Network as a Service (NaaS) PlayBook

The NaaS PlayBook is the vehicle to serve as the planning and implementation guide for RST NaaS initiatives.

The PlayBook aspires to…

  • Aid NaaS operators in the planning, deployment and operation in the rural environment.
  • Provide a vehicle to serve as the planning and implementation guide for NaaS initiatives.
  • Create a multi-user engagement “product” where NaaS operators seek guidance, instruction and decision support.

The PlayBook will…

  • Clarify strategic plans; consider and confirm optimal Network Architectures.
  • Provide functional overviews, process definition and best practices.
  • Engage the user to explore.
  • Serve as a “network development guide” sequenced with a high level planning effort that in turn drives network design, implementation and management.
  • Allow users to “opt-in” and interact from planning through implementation.


The NaaS PlayBook is the vehicle to serve as the planning and implementation guide for RST NaaS initiatives. It consists of Frontend and Backend streams comprised of modules, atomic units built around specific objectives and select content and complemented by methods to encourage engagement relevancy.

 

 

1. Strategy and Design

A Strategy & Scope Definition plays a critical role for NaaS operators as it defines the strategic goals as well as the strategic decisions to achieve them.

The Strategy & Scope definition is fundamental in the NaaS operator initiative as it provides a sense of direction and outlines measurable goals. Through an optimal definition of these aspects, the leadership team can ensure alignment on the scope and priorities towards fulfilment of the strategic objectives.

This module guides the NaaS operator through an overview of the Strategy & Scope fundamentals and the imperative of its optimal definition and implementation. Furthermore, it provides an Implementation Process that can be tailored to translate the NaaS operator mission and vision into strategic objectives that can be tracked.

The main outputs of this module are the Strategies Definition for critical aspects of the Naas operator. These strategies direct the decisions on subsequent modules in order to align them with the high-level objectives. Complementary outputs are, among others, the Mission & Scope Definition and the Tracking Methodologies to evaluate the implementation of the defined strategies.

1.1 Module Objectives

This module will enable the NaaS Operator to define the Strategy & Scope of the NaaS initiative. The specific objectives of this module are to:

  1. Provide procedures to enable NaaS operator to perform a self-assessment analysis based on an examination of the particular environment and market.
  2. Guide the NaaS operator to clarify the high-level scope by defining its own tailored mission and service definition.
  3. Instruct the NaaS operator in the strategy and long-term objectives definition to guide decisions in downstream activities.
  4. Provide methodologies to control and manage the defined strategies implementation and updates.

1.2 Module Framework

The Module Framework in Figure 1 describes the structure, interactions and dependencies among different NaaS operator areas.

Strategic Plan & Scope is the first stream to build the NaaS operator. The modules in this stream aim to drive strategic decisions that impact all the downstream streams and modules. From this point, High Level Network Architecture can be defined establishing the guidelines for Network Design which in turn kick-starts the deployment phase.

Figure 1. Module Framework

Figure 2 presents the Strategy & Scope process view where the main functional components are exhibited. Critical module inputs are further described and examined.

Figure 2. Module Process View

The rest of the module is divided into five sections. Section 2 is a birds-eye view of the Strategy & Scope Fundamentals as well as a high-level view of their development process. Section 3 focuses on the procedures to examine the particular environment and market. Section 4 describes processes to define the NaaS operators mission and scope. Then in section 5, methodologies to generate the overall strategies and long-term objectives are provided. Finally, Section 6 includes procedures to control and manage the implementation and updates of the defined strategies.

 

2 Strategy & Scope Fundamentals

This section presents a general overview of the baseline concepts to develop the NaaS Initiative Strategy & Scope.

2.1 Strategy & Scope Imperative

The Strategy & Scope of the NaaS operator is the process of an organization deciding their corporate direction, objectives and priorities, and then aligning their resources to accomplish the actions necessary to meet them.

The main outcome of the Strategy & Scope module is the Strategy definition around certain critical areas within the NaaS operator. The Strategies will guide the decision of subsequent phases (e.g. Network Design and Deployment). This means that the Network Design and Deployment streams use the strategic decisions of this module to generate its own processes. For this reason, Strategy & Scope is the first step to build the NaaS operator.

Additional outputs from this module are Organization Strategies and Long-Term Strategic Objectives. These outputs are the starting point and provide a general guide for the High-Level Project Plan, which aims to answer how the Strategy and Objectives will be achieved and who must do what to accomplish them.

Finally, as stated before, the Strategy & Scope module establishes the direction of the NaaS operator, including the financial resources, so it maintains a relationship with the Financial Plan. In other words, the Financial Plan outlines the allocation of financial resources to reach the broad goals set out in strategic planning. However, the Financial Plan is out of the scope of this module.

2.2 Process Overview

Figure 3 displays the process to develop the Strategy & Scope Definition for NaaS operators in a logical and well-structured sequence. It offers a roadmap for functional leaders that must identify and prioritize initiatives at a functional level.

Figure 3. Strategy & Scope Definition Process

The first step, Research & Analysis, provides the methodologies to analyze the conditions of different environments that the NaaS operator is subject to. This information can be collected via web research on the operating environment and telecom-industry environment reports. The survey application on customers is also a common practice. The primary outcome of this process is the SWOT analysis.

The next step, Mission and Scope, addresses the what (mission) and where does the NaaS operator want to be in the long term (vision) of the NaaS operator. The definition of these elements is an essential part to building the strategic foundation and developing a strategy. Based on these elements, the competitive advantages can be identified. An additional step for the NaaS operator, is to define the technologies and services that will be part of its catalog.

The Strategy Design step contains the methodologies to develop the strategies around different NaaS operator areas. These strategies guide the decisions on subsequent modules. Furthermore, it provides the methods to transform these strategies into realistic goals.

Finally, the Governance step provides guidance to move a strategic plan from a document that sits on the shelf to actions that drive organizational growth.

3 Research & Analysis

This section aims to describe procedures to examine the particular NaaS operator environment and market.

3.1 Strategic Issues & Risk Identification

Strategic issues are critical unknowns that need to be resolved by the strategic planning process. These issues can be problems, opportunities, market shifts or anything else requiring special consideration by the NaaS operator.

Strategic issues are developed and identified based on input from the leadership team. These issues should be a summary of critical topics that need to be addressed during the planning process. The idea is to call these issues out early, so the important areas are not lost among a lot of data, details, and ideas.

In the NaaS operator context, some strategic issues can be:

  • How to fund the NaaS operator initiative?
  • What are the options to acquire the radio spectrum?
  • What does the organization look like?

After identifying the Strategic Issues, they need to be recorded into the Strategy & Scope Definition Template as they will be used for the Strategies definition.

3.2 Operating Environmental Scan

The environmental conditions have a significant impact on the direction of the NaaS operator. Since the environment is prone to continuous changes, the organizations that are not aware of their business environments are likely to fail.

The environmental scan is the process of identifying, collecting and evaluating information about elements that are external influences on the NaaS operator. This process helps in a greater understanding of environmental changes and guides decisions on the NaaS operators strategic direction.

The environmental factors impacting the strategic planning can be grouped into two broad categories: the Operating Environment and the Telecom Industry Environment as shown in Figure 4. This section focuses on analyzing the distinct aspects regarding the Operational Environment.

Figure 4. Environment Factors Diagram

More details regarding the specific elements within each category in Figure 4 are provided in the following subsections.

The operating environment, which indirectly affects the NaaS operator comprises Political, Economic, Social and Technological factors. Thus, the operating environment scan is also named as PEST analysis and is part of the Strategy & Scope process. The result of the Operating Environment Scan must be recorded into the Strategy & Scope Definition Template as it will be used on the SWOT Analysis.

Political (includes legal and regulatory)

Includes the political-legal factors comprising the countrys laws and regulations that can influence the NaaS operator objectives and operations.

  • National Spectrum Regulation: The local Telecommunications Regulatory Entity establishes the opportunities for the spectrum to be used by the NaaS operator.
  • National trade agreements: Nowadays, there is a greater tendency towards international and regional trade agreements, which enables the access of different vendors and suppliers of varied origins to a common marketplace. Privatization and Deregulation: The NaaS operator must identify and evaluate the level of government control in the telecommunication market as it can impose some restrictions and limitations. Legal Compliance: Apart from the common labor and employment laws, there might potentially exist other laws and licenses that are essential to be complied with (e.g. telemarketing and privacy laws). National Security: Another potential source of political constraints for the telecom sector is the measurements that the local government imposes to monitor and control communications. This could impose additional requirements in the architecture definition.

Economic

The role of economic factors is important in the context of the telecom industry. NaaS operators should account for the national economy, including inflation rates, interest rates, and most importantly, disposable income.

  • Interest rates, inflation and taxes: These metrics affect the overall NaaS operators operation directly and, therefore, the pricing per plan offered to customers. Furthermore, the inflation and currency rates coud impact the Capital Expenditure (CapEx) if certain equipment is imported.
  • Income levels: The demand for telecommunications services such as mobile phones and Internet can be influenced by disposable income levels.

Social

Social factors have a profound influence on the telecommunications industry and its profitability. This is because it has a direct impact on the number and type of services demanded by users.

  • Demographics: Demographic factors (e.g. age, gender, family size) helps to identify consumer behavior and determine the demand for products and services.
  • Lifestyle: The use of Internet-based services has expanded as a higher number of people are using social media (e.g. Facebook, YouTube). These services require good connectivity that can only be satisfied with 4G coverage. Furthermore, the availability of 4G mobile devices (i.e. device penetration) must be identified.

Technological

The telecommunications industry is highly dependent on technology changes. The NaaS operator should gain most of these technological trends.

  • New Telecom Trends: NaaS operators must be aware of raising technologies that can potentially leverage its planning and operation. For instance, disaggregated hardware or cloud-based solutions can reduce the expenditures on the network deployment. Moreover, 5G radio technology can potentially increase the transmission rates.
  • End-devices Evolution: The emergence of advanced end-devices such as tablets and smartphones, drive the requirement of 4G connectivity. Moreover, several other devices will utilize a connection like biometrics, whose use on smartphones is expected to grow. 

3.3 Telecom Industry Environment Scan

As shown in Figure 4, the NaaS operator must be aware of the Telecom Industry Environment. This section provides general guidelines to collect information about customers, competitors, and other market variables to generate insights for strategies definition.

3.3.1 Telecom Conditions Evaluation

Suppliers Environment

A part of the Environmental Analysis is the evaluation of the available suppliers and their effect on the environment. The NaaS operator suppliers environment is mainly composed of the following elements:

  • Network equipment providers: The available network equipment providers and the cost to the NaaS operator must be analyzed. Furthermore, it is fundamental that the examined vendors offer Proven Technology and provide evidence of their equipment in production. Some of the providers will require international transportation to the NaaS operator region, which may potentially increase the cost and delivery time.
  • Service Providers: The different service providers with coverage in the NaaS operator region must be examined. These providers include among others Transport Providers and Fiber / Tower Companies.
  • Workforce, or suppliers of labor: This element is affected by the availability of a qualified and experienced telecommunications sector workforce and also by the consolidation in the regional labor market.

The information regarding suppliers environment can be collected by performing a research on the web sites of the different suppliers.

Geographic Conditions

An analysis regarding the terrain conditions must be conducted, especially in rural areas. The specific geographic features (e.g. mountains, dense forest) may significantly complicate the network deployment for remote populations as the access to the site for construction and maintenance may complicate. Furthermore, harsh climates (e.g. extreme hot or cold) can make the operation and maintenance tasks of the site more challenging.

This analysis has a direct impact on the subsequent design and deployment phases as often extensive microwave or fiber optic laying must be required in the transport network. If the terrestrial solutions are not feasible, NaaS operators often have to rely on satellite transport with additional expensive bandwidth fees to connect remote areas.

In addition to the terrain, the climate can also have a significant impact on the satellite signals as these are vulnerable to rain conditions that can occur regularly in tropical areas.

In further deployment phases, it is important to perform a Technical Site Survey (TSS) on each site to clearly define the geographic conditions.

Basic Infrastructure

The network deployment in rural and remote locations is directly impacted by a lack of basic infrastructure (e.g. power grid, road access). If adverse conditions are presented, the Naas operator must build each site in a self-sufficient manner (e.g. solar energy-based). This represents an increment in the deployment costs and subsequent operations and maintenance costs.

Furthermore, the quality of the basic infrastructure must be identified as bad quality infrastructure may represents additional costs on the site operations (e.g. bad quality power grid may require backup batteries).

The Technical Site Survey (TSS) on each site in further deployment phases clearly defines the existing basic infrastructure conditions.

3.3.2 Customer Segmentation and Targeting

The NaaS operator should recognize that all customers have to be approached in different ways as their needs and values are different. Rather than trying to compete in the entire market, the NaaS operator must identify parts of the market that can be served best and most profitably.

The most effective way to achieve this is to segment customers. The goal of creating customer segments is so that specific customers that have similar needs can be targeted with the same products and pricing.

The following questions can help to identify the target Customer Segments:

  • What is the biggest customer segment?
  • Do customers fall into any logical segments based on needs, or characteristics?
  • Are there one or two characteristics that will help to segment customers? (Geography, price, need, etc.)

In the context of NaaS operator environment, the potential customers are:

  • Mobile Network Operators (MNOs): An MNO with no 4G coverage in the desired area using its own granted spectrum or via roaming agreements and the network infrastructure of the NaaS operator.
  • Mobile Virtual Network Operator (MVNOs): Virtual operators that can offer mobile services by using the network services of the NaaS operator.
  • Wireless Internet service provider (WISP): These operators can extend its coverage area by using its own granted spectrum and the network infrastructure of the NaaS operator.
  • End-users: These are end-users subscribers that directly buy LTE services from the NaaS operator.

Once segments are identified, the NaaS operator must determine its particular needs and the most appropriate way to satisfy them by providing value. The following process illustrates the process to perform customer segmentation.

  1. Segment Name: Establish a name for the target customer segment.
  2. Evaluate: After the customer segments have been identified, evaluate them based on the following criteria:
  • Can the size of the segment be measured?
  • Can this segment be reached through clear communication channels?
  • Are there enough customers in the segment to make it profitable?
  • Will this segment respond differently to product and service offerings than other segments identified?

  • Identify Customer Needs: List the needs or wants of this customer segment. Target customer segments can be grouped based on similar needs, motivation, or behavior.
  • Identify Customer Characteristics: List the characteristics that describe this segment. The major categories of characteristics are: demographic, geographic and usage.
  • Customer Profile Development: Create a customer profile that uniquely describes this target customer segment.

 

In the context of NaaS operator environment, some alternatives of customer segment definition are:

  • Per consumption level: stratified per consumption unit (traffic volumes, number of subscribers)
  • Per geographic zone: stratified per geographic area.

3.3.3 Competitive Analysis

The competitive analysis aims to assess the opportunities and threats that may occur from those organizations competing for the same business. The NaaS operator needs to have an understanding of what competitors are or arent offering the same service to potential customers.

Unless explicitly stated, along this document the term customer refers to entities that buy services directly from the NaaS operator (e.g. MNOs, MVNOs). In contrast, end-user term refers to subscribers that use the LTE services. It is important to mention that not all the end-users buy services directly from the NaaS operator, as some of them will buy the service from an MVNO.

In order to identify the organizations that compete for the same customers, the following actions must be performed:

  • Make a list of similar organizations by name or group. What are each ones strengths and weaknesses?
  • Are these opportunities or threats to the NaaS operator?
  • What are the competitive advantages of similar organizations?
  • What is happening with these similar organizations? Are they growing or shrinking?

The following subsections examine different types of organizations that compete for the same group of customers as NaaS operators. They can be divided into two main groups: Direct competitors and Substitute Products.

Direct Competitors

In the NaaS operator context, the direct competitors are:

  • Mobile Network Operators: An MNO with a well-established existing network represents a competitor for the NaaS operator. An MNO is also a direct competitor when rural areas are part of their expansion plan.
  • Wireless Internet service provider (WISP): These operators provide Internet services at designated hot spots (access points) using a wireless connection such as Wi-Fi or provide fixed wireless to end customers homes and business.

Substitute Products

A substitute product performs the same or a similar function as the NaaS operator offer but by different means. Substitutes offer the greatest threat when they can provide customers with better service at lower costs through changes that improve the value of their products or services.

In the NaaS operator context, the main substitute products can be in the form of:

  • Wireline providers: Traditional wireline providers that offer Internet and voice service via cable and digital subscriber lines (DSL). However, these types of providers do not have great penetration in rural areas.
  • Satellite internet providers: Providers that offer Internet access via satellite link. Although these operators provide great coverage in rural areas, the high installation and operational costs translate into expensive user fees.

In order to quantify the specific threat of substitute products, it is essential to look at the price-performance trade-off between the substitute and the NaaS operator services.

3.4 SWOT Analysis

This section presents a Methodology to develop a SWOT (Strengths, Weaknesses, Opportunities, and Threats) Analysis to evaluate the NaaS operators strategic position.

3.4.1 Identify Opportunities and Threats

Opportunities are situations that potentially exist but must be acted on so to benefit the NaaS operator. Most relevant opportunities are those that offer important avenues for profitable growth, those where a company has the most potential for competitive development, and those that match up well with the financial and organizational resource capabilities that the company already possesses or can acquire.

In contrast, Threats refer to external conditions or barriers that may prevent the NaaS operator from reaching its objectives.

The areas presented in Table 1 can help to identify opportunities and threats that can be identified using the results of the Operating and Telecom Industry Environmental Scan.

Area

Potential opportunities and threats

Operating Environment

– Political/Legal – Environment – Social – Technological

Telecom Industry

– New competitors – Substitute products – Competitive rivalry

Competitors

– Identified competitors – Strengths, weaknesses

Table 1. Areas of potential Opportunities and Threats

3.4.2 Identify Strengths and Weaknesses

Strengths refer to the areas (e.g., operations, management, marketing) in which the NaaS operator performs successfully. An analysis of organization strengths should be market and customer-focused because strengths are only meaningful when they assist in meeting customer needs. Strengths give the organization enhanced competitiveness.

On the other hand, Weaknesses refer to any limitations that the NaaS operator faces in developing or implementing a Strategy. A weakness is something an organization lacks or does poorly in comparison to others, or a condition that puts it at a disadvantage. Weaknesses should also be examined from a customer perspective because customers often perceive weaknesses that an organization cannot see.

The strengths and weaknesses can be discovered via brainstorming process among the elements of the leadership team and by breaking them down in the areas presented in Table 2.

Area

Potential strengths and weaknesses

Capabilities

– Human – Organizational – Knowledge

Resources

– Financial – Physical – Intangible

Processes

– Deployment – Operational – Customer management

Table 2. Areas of potential Strengths and Weaknesses.

3.4.3 SWOT Development

A SWOT analysis is a methodology to examine an organization by looking at the internal strengths and weaknesses in relation to external opportunities and threats. This analysis takes as an input all the previous analysis (i.e. Operating and Telecom Industry Environmental Scan, Customer Segmentation, Competitive Analysis). By creating a SWOT analysis, all the important factors affecting the organization can be seen together in one place. Its easy to read, easy to communicate, and easy to create.

The following list considers some questions to be asked during the development of the SWOT. These questions should have been already answered in the analysis prior to the SWOT development:

  • Strengths: What do you do well (in sales, marketing, operations, management)? What are your assets? What are your core competencies? What differentiates you from your competitors? Why do customers want to buy from you?
  • Weaknesses: Where do you lack resources? What can you do better? What necessary expertise do you currently lack? In what areas do your competitors have an edge?
  • Opportunities: What new needs of customers could you meet? What are the economic trends that benefit you? What are the emerging political and social opportunities? What are the technological breakthroughs? What niches have your competitors missed?
  • Threats: Where are the red alerts in your environment? What are the negative economic trends? What are the negative political and social trends? Where are you vulnerable?

Once the SWOT Analysis is completed, the information should be used to start the development of strategies that will leverage the NaaS operators strengths to pursue opportunities, while also countering identified weaknesses and threats that might undermine its efforts. The SWOT analysis is a crucial input to develop strategic alternatives and potential goals.

Additionally, the Naas operator can use the SWOT Analysis Template to support the SWOT Analysis development.

4 Mission and Scope

The content of this section illustrates the procedures to establish a charter by defining the particular mission, vision and service definition for the NaaS Initiative.

4.1 Mission Statement Development

The mission statement is a declaration of the NaaS operators purpose and spotlights the customer needs that are endeavoring to be met. To build a solid foundation for a successful organization, it is essential to have a written, clear, concise, and consistent mission statement that simply explains the NaaS operators purpose.

The mission statement should serve as a guide for day-to-day operations and as the foundation for future decision-making. The following guidelines should be considered when writing or evaluating the mission statement:

  • Focuses on satisfying customer needs: a mission statement should focus on satisfying customer needs rather than on a program or service. It states who are the customers, what needs the NaaS operators aim to satisfy and how those needs are satisfied.
  • Based on core competencies: NaaS operators should base its mission on a competitively superior internal strength, unique capability or resource that the organization performs well in comparison to similar operators.
  • Realistic: The mission statement should be realistic, avoiding making the mission too narrow or too broad.
  • Specific, short, sharply focused and memorable: It should be a precise statement of purpose that describes the essence of the organization in words that customers and stakeholders can remember.
  • Clear and easily understood: Develop and write the mission statement in a way that it can effectively express the NaaS operator purpose.

Additionally, the Naas operator can use the Mission Statement Assessment Checklist to assess the above criteria on the developed mission statement.

4.2 Vision Statement Development

The vision statement provides a clear mental picture of what will the NaaS operator look like in 5 to 10 years from now. Forming a strategic vision should provide long-term direction, delineate the organizational activities to be pursued and the capabilities the NaaS operator plans to develop. It serves as a unifying focal point for everyone in the Organization.

An effective vision statement consists of the following elements. The specific vision statement generated by the NaaS operator may nor incorporate all these elements, but they must be considered when casting the statement:

  • Future casting: Provides a picture of what will the NaaS operator look like in the future.
  • Purpose-Driven: It should be appropriately worded to give internal staff a sense of purpose within the NaaS operator.
  • Capitalizes on Core competencies: It builds on strengths, and unique capabilities, resources and assets.

4.3 Identify Competitive Advantages

The competitive advantage is what the NaaS operator does or potentially could do better than similar organizations. One result of a well-developed and executed strategic plan is to develop a unique competitive advantage. A sustainable competitive advantage(s) is the foundation, the cornerstone of the strategy.

The following criteria must be considered to evaluate the Competitive Advantage identification:

  • Consistent difference: Customers must perceive a consistent difference between the NaaS operator services and those of its competitors. This difference needs to be obvious to customers, and it must influence their purchasing decision.
  • Difficult to imitate: A good competitive advantage must be difficult to imitate. This might be in the form of people, proprietary knowledge within the organization, or business processes that are behind the scenes.
  • Constantly improved: The first two bulleted items above must create activities that can be constantly nurtured and improved upon in order to maintain an edge over the competition.

Some examples of competitive advantage in the NaaS operator context are:

  • Be the only operator which offers mobile broadband data services in the region.
  • Offer LTE services at a lower cost than the possible investment on the construction and operation of a new mobile network.

4.4 Technologies and Services Definition

This section examines different technologies and services that can be included as part of the catalog to be offered to customers.

4.4.1 Radio Technologies

4G Long-Term Evolution (LTE)

4G LTE is a wireless communications standard developed by the 3rd Generation Partnership Project (3GPP) and it is reflective of the coverage consumers now expect. LTE delivers the ability to use end devices to their peak functionality.

The recommendation is that LTE must always be the default technology to be deployed by NaaS operators.

Pre-Existing 3G (UMTS) / 2G (GSM) Network

The rural areas may have coverage with legacy technologies (e.g. 3G UMTS and 2G GSM) provided by some existing MNOs. However, these technologies are not capable of supporting todays user requirements.

It is not recommended to deploy 3G equipment as the use of this technology will be declining in the coming years. However, in case of deployment of 3G equipment, the NaaS operator must ensure that it can be depreciated while still being used.

The recommendation is that 3G and 2G technologies should not be deployed and only be considered for CS Fallback capabilities.

5G Radio Technology

The fifth-generation (5G) of cellular technology increases the speed and responsiveness of wireless networks. Over this network, data can be transmitted at multi-gigabit speeds, with potential peaks of 20Gbps. Moreover, it offers a latency of 1ms that enable real-time applications. 5G will also increase the available bandwidth due to advance antenna technology implementation.

Despite all the advertised capabilities, 5G deployments have not been massively implemented as the technology and equipment standards are still in process of development.

The recommendation is that 5G radio technology should not be deployed in the NaaS operator network until successful tests are successfully completed in NaaS operator laboratories or flagship rural deployments.

4.4.2 Voice Services

The NaaS operator must evaluate if Voice Services will be part of its catalog of services as it represents special considerations in the network architecture. It can select to only provide data services for their customers.

If the NaaS operator wants to offer voice services as part of its catalog service, there are three available options that are described in the following subsections. Further details on the implementation of these alternatives are provided in the Mobile Core Architecture module.

Circuit Switched FallBack (CSFB)

In this scenario, the LTE handset falls back to 3G or 2G to establish a traditional circuit switching connection. As CSFB is not strictly an LTE based voice service, rather EPC assists the device to fall back to a 2G/3G network. It presents many drawbacks such as, higher call-setup-delay or just acceptable quality voice.

However, if the NaaS operator has a pre-existing 2G/3G network then the CSFB functionality must always be deployed for simplicity.

Voice over internet protocol (VoIP) / Over the top (OTT) Application

VoIP is the transmission of voice and multimedia content over Internet Protocol (IP) networks. This service is also referred as Over-the-top (OTT) Application and PoC (Push-to-Talk over Cellular). These type of applications can be free (e.g. WhatsApp, Facebook) or be provided by the VoIP service provider.

VoIP is the recommended alternative when the NaaS operator wants to provide voice services through its 4G-only network and has budget limitations. However, it should be noted that the voice quality will never match VoLTE.

Voice over LTE (VoLTE)

VoLTE is the industry-recognized solution for providing a packet voice service over IP via LTE access technology. When the NaaS operator has the economical/knowledge capabilities and agreements with an IMS network provider, then the option that will deliver HD voice is VoLTE.

Furthermore, NaaS operators should consider the availability of VoLTE-capable devices on its market and the feasibility of VoLTE business case with its customers. Additionally, the customer (e.g. MNO) need to support VoLTE.

5 Strategy Design

The present section describes the definition of the strategy and long-term objectives that ultimately will guide the downstream activities.

5.1 Organization-Wide Strategies Definition

An organization-wide strategy is a general statement(s) that guides and covers a set of activities. It answers the question how. By consistently executing an organization-wide strategy, the NaaS operator can provide a product or service thats better than its competition.

In this step of the planning process, the synthesized information in the SWOT analysis, the mission, vision and competitive advantages are used to develop the organization-wide strategy or strategies that will guide goal and action planning.

The following elements summarize the core of strategy and should be accomplished in the definition of an organization-wide strategy:

  • Diagnosis of the situation: A good strategic diagnosis does more than explain the situation; it also defines the domain of action. This can be accomplished during the SWOT process.
  • Guide decisions: A good strategy directs and constrains action without fully defining its containment. It tackles the obstacles identified in the diagnosis by creating or drawing upon sources of advantage. This is the intended outcome during this phase of the planning process.
  • Design of coherent, coordinated action: Goals and actions at various levels of the organization are developed in concert with the guiding strategy. These elements are developed in the subsequent phases.

There are certain areas within the NaaS operator that require the definition of a strategy that guides the decision of subsequent phases. These areas are presented in Figure 5 and are examined in the following subsections.

Figure 5. Strategy Map

Additionally, the Naas operator can use the Strategy & Scope Definition Template to support the definition of their own strategies.

5.1.1 Funding Strategy

A primary element to develop a telecommunication network is to identify the funding methods to deploy the network. There are different strategies for funding telecommunication projects that are presented in Figure 6. NaaS operators must evaluate which of them will be used to fund the initiative. The final decision must consider, among other elements, the applicable political and regulatory conditions and it must be recorded in the Strategy & Scope Definition Template.

Figure 6. Funding Strategy Alternatives

In the following subsections, multiple alternatives that can be used in different countries around the world are presented.

Self-financing

Most of the time, self-financing method is only applicable when the NaaS operator already has an existing network in the country. In this way, the costs of infrastructure can be reduced as existing towers and fiber can be used to support the deployment of the 4G network. Furthermore, the profits from the existing network can be used to build out the new network deployment.

In case that the cost of deployment exceeds the NaaS operators financial capabilities, an alternative method must be evaluated.

Telecom subsidies

An alternative option is to apply for telecom subsidies in the region. A definition of a subsidy is a grant or a monetary gift, usually given from the government to a specific project. Another form of subsidies is that the government reduces the taxes for a company that needs assistance; thereby they get more financial resources.

Joint venture loans

Joint venture loans are an option when the loan for the investment cannot be covered for one single company. This means that two or more companies share the risks and benefits of the loan.

Joint ventures can result in an effective partnership, but to make this happen, they must be structured. To ensure this, it is recommended that an independent president is chosen.

A joint venture often ends when the predetermined goal has been reached (e.g. the deployment completion) so all involved parties need to be prepared for that.

Public-Private Partnership

This partnership means that the private and public sectors share the risk for the project. If there is any risk that they together cannot control, the public sector usually takes responsibility for that risk.

Principally a public-private partnership is a form of privatization, but with a big difference that the public sector is also involved. The partnership is involving the government and one or more private sector companies.

5.1.2 Operator Engagement Strategy

The traditional approach of network deployment, where each operator builds its own network infrastructure is not efficient in rural markets with low population density. In turn, the NaaS operator must analyze ways to co-operate with other operators on expanding network coverage whilst preserving healthy competition in service provision. In this way, an MNO can become a partner for the NaaS operator.

The Operator Engagement has multiple dimensions in the sense that include elements of the Spectrum, Commercial and Infrastructure Strategies. These elements are examined in its corresponding section, but the final definitions must be consolidated in an Operator Engagement Package in order to present them to the operators.

Figure 7 illustrates different alternatives to perform the Operator Engagement. The NaaS operator must evaluate them and select the most appropriate according to its conditions. The final decision must be recorded in the Strategy & Scope Definition Template.

Figure 7. Operator Engagement Strategy Alternatives

The following subsections describe the way to engage different types of operators as well as their impact on the NaaS operator strategy. It must be noted that the decision of the operators to be considered to interconnect with, has a direct impact on the RAN and Mobile Core implementation.

5.1.2.1 Mobile Network Operators (MNO)

The primary driver that an existing Mobile Network Operator can consider to set an agreement with the NaaS operator is to extend its coverage to rural scenarios. For this, the scenarios in the following subsections can be implemented.

The impact on the NaaS operators Radio Access Network and Mobile Core implementation are further detailed in the respective Architecture Modules.

Network Roaming Agreements

In this scenario, NaaS Operator owns the RAN and the core network and redirects the traffic to MNOs core networks. Additionally, the user might have to enable roaming on their phone to access the NaaS network.

This scenario for the NaaS operator implies that a complete RAN and transport infrastructure must be deployed using its own spectrum. Additionally, the Mobile Core must also be implemented along with the interconnection to MNOs Mobile Core.

RAN Sharing Agreements

NaaS Operator can provide its RAN infrastructure to multiple MNOs based on different RAN Sharing schemes (e.g. Multi-Operator RAN, Multi-Operator Core Network). In this scenario, a pool of radio spectrum is formed using the available sources (e.g. NaaS operator, MNO). In addition, MNOs rely on the NaaS radio/transport infrastructure to provide service to the end user.

This scenario for the NaaS operator implies that a complete RAN and transport infrastructure must be deployed and its radio spectrum (if existing) can be used to conform the pool of spectrum.

5.1.2.2 Mobile Virtual Network Operator (MVNO)

A Mobile Virtual Network Operator (MVNO) is a virtual operator that offers mobile services under its trademark by using network services from a third-party mobile operator. The MVNOs can be further classified in different categories according to their characteristics.

The impact on the NaaS operators Mobile Core implementation are further detailed in the Architecture Module.

Full-MVNO

In this scenario, the NaaS operator just provides the access network infrastructure and, sometimes, part of the core network. Therefore, the MVNO must build and maintain its own Mobile Core. As a result, the MVNO manages its own customer base and associated services independently from the host operator.

This scenario for the NaaS operator implies that a complete RAN and transport infrastructure must be deployed. In addition, the NaaS operator must possess granted radio spectrum. Finally, the interconnection to MNOs Mobile Core needs to be implemented.

Light-MVNO or Re-seller

The MVNO just provides its brand and its distribution channels. In contrast, the NaaS operator provides the access and transport, as well as the mobile service. This is the model that requires the lowest investment from the MVNO, therefore the fastest to implement.

A Light-MVNO delegates the networks operational management to its host operator, in order to focus on customer relations.

This scenario for the NaaS operator implies that a complete RAN and transport infrastructure must be deployed. . In addition, the NaaS operator must possess granted radio spectrum. Finally, a complete Mobile Core must also be implemented.

5.1.3 Spectrum Strategy

Efficient spectrum management has an important impact on the quality and reach of affordable mobile services. Figure 8 displays the different alternatives to obtain and manage the radio spectrum.The NaaS operator must evaluate each alternative and select the most appropriate according to its conditions. The final decision must be recorded in the Strategy & Scope Definition Template.

Figure 8. Spectrum Strategy Alternatives

The following subsections examine different approaches to obtain and manage the radio spectrum.

It should be noted that a spectrum license should also be considered for transport microwave equipment. However, microwave spectrum considerations are deeply analyzed in the Transport & IP Architecture module.

5.1.3.1 Own Spectrum

The first alternative is that the NaaS operator manages and operates its own granted spectrum. The following subsections examine different alternatives to achieve this.

Re-farming

If the NaaS operator has an existing operating mobile network, it can use the already licensed spectrum to deploy LTE services. Existing low-frequency spectrum (e.g. 900MHz) used for 2G/3G services can be more effectively reused for 4G service deployment.

Spectrum re-farming is a cost-effective way to increase capacity in existing networks without the need to bid for new spectrum and by reutilizing infrastructure to overlay LTE.

Standard Licensed spectrum

The NaaS operator can follow the standard licensed mechanisms to acquire sufficient spectrum to provide broadband mobile services.

However, the cost of the spectrum fee may impact the overall business case. Additionally, the process of acquiring spectrum is highly time-consuming given its complexity.

Rural Licensed spectrum

Spectrum regulators have been encouraged to clear the 700 MHz and 800 MHz spectrum through the migration to digital television, which can release the so-called digital dividend spectrum for use by mobile services, especially in rural areas.

Furthermore, in some countries, the local government provides specific telecom subsidies and grants to these types of radio bands in order to promote the coverage expansion in rural areas. The NaaS operator can analyze its applicability to these special programs.

Unlicensed Spectrum

The utilization of unlicensed spectrum in the 5GHz range allows an operator to deploy LTE services without the necessity of a license for spectrum. However, the use of unlicensed spectrum imposes transmission power limitations, which limits the coverage area.

In addition, the NaaS Operator needs to verify if other operations are using unlicensed spectrum in the area, as this may cause a considerable amount of interference.

The NaaS Operator must be aware that the operation of LTE in unlicensed spectrum may not be supported by all user devices.

5.1.3.2 3rd Party Spectrum

NaaS operators can use a 3rd party spectrum assigned to provide LTE services. The following subsections examine different options to accomplish this.

MNO Spectrum

NaaS Operators can use the MNO granted spectrum to provide LTE services via different RAN Sharing schemes (e.g. Multi-Operator RAN, Multi-Operator Core Network). In this scenario, the spectrum of the MNO is used and, in some cases, shared.

This scenario imposes some special considerations on the overall network architecture that need to be taken into account and are further explored in the Transport and IP Architecture Module.

Spectrum Leasing

NaaS operators can lease the spectrum from a third party (e.g. MNO, WISP) in order to provide LTE services. This represents an additional recurrent cost to NaaS operators that may impact the overall deployment business case.

In addition, an accurate capacity forecast analysis must be performed in order to determine the amount of bandwidth to lease. If this is not well-forecasted, the NaaS operator may over or underutilize the leased bandwidth.

5.1.4 Commercial Strategy

The commercial strategy of the NaaS operator is composed of two elements: Revenue Strategy and Marketing Strategy. Figure 9 displays the different alternatives related to the Commercial Strategy. The NaaS operator must evaluate each alternative and select the most appropriate according to its conditions. The final decision must be recorded in the Strategy & Scope Definition Template.

Figure 9. Commercial Strategy Alternatives

5.1.4.1 Revenue

The Revenue Strategy must identify Revenue Streams within the NaaS operator environment. The nature of each of the streams is different, so a specific Pricing Strategies are required for each of them.

The applicable pricing strategies in the context of the NaaS operator are:

  • Fixed Fee: In this case, the NaaS operator charges a recurrent fixed fee for providing the services to the 3rd party independently of the service consumption. The recommended frequency to perform this activity is on a monthly basis.
  • Traffic-based: The NaaS operator can bill according to the generated traffic by the 3rd party subscribers.
  • Subscribers-based: Alternatively, in the Light-MVNO scenario, the charged fee can vary according to the number of subscribers to be supported by the NaaS operator network independently on the specific user behavior.

In this way, the different Revenue Streams can be classified accordingly. Table 3 displays a typical Revenue Streams classification for a NaaS Operator.

Revenue Stream

Description

Customer

Pricing Strategy

LTE Service Provisioning

Revenue is generated by the provision of the LTE service to 3rd party entities.

MVNO, specially Light-MVNOs

– Fixed Fee – Traffic-based – Subscribers-based

RAN Infrastructure Leasing

Revenue generated by leasing its RAN infrastructure

MNOs or WISPs

– Fixed Fee

Roaming Agreements

Revenue generated for providing LTE service to MNO’s subscribers in an area where only the NaaS operator have coverage

MNOs

– Traffic-based

Table 3. Typical Revenue Classification for a NaaS Operator.

5.1.4.2 Marketing

The Marketing Strategy describes the specific types of Marketing Models that NaaS operators will implement in order to promote its services among its direct customers and end-users.

The marketing models that the NaaS operator should consider are presented in the following subsections.

Customer-centered

This model is part of the Engagement Operator package and is focused on promoting the differentiators of the NaaS operator services among its direct customers (e.g. MNOs, MVNOs). The aim is to demonstrate and explain the advantages of using NaaS operator services (e.g offer cheaper broadband mobile service focused on rural connectivity).

Once an engagement is established, periodic approaches can be performed in order to consolidate and strengthen NaaS operator position.

The usual channels of communication used by this model are Email and Face-to-face Meetings.

End-user-centered

The focus of this model is to announce and promote the NaaS operators services among end users (i.e. mobile subscribers). These end users may not buy directly from the NaaS operator, however, they need to be aware that they are making use of its services. This type of marketing is very important when a NaaS operator brings connectivity to a new area, so the inhabitants aware that this new connectivity exists.

This model expects to encourage further positive purchase-related behavior. This can be in the form of subscribers visiting the NaaS operator website, read about it, and share.

The typical channel communication used by this model is mass media (e.g. Social media, TV spots, radio advertisement, printed newsletter)

5.1.5 Deployment Strategy

The commercial strategy of the NaaS operator comprises two aspects: the Logistics & Warehousing and the Infrastructure strategy. Figure 10 displays the different alternatives related to the Logistics & Warehousing Strategy. The NaaS operator must evaluate each alternative and select the most appropriate according to its conditions. The final decision must be recorded in the Strategy & Scope Definition Template.

Figure 10. Deployment Strategy Alternatives

5.1.5.1 Logistics & Warehousing

Logistic & Warehousing describes the form in which all the network equipment and ancillaries are stored and distributed. The possible alternatives to implement these services are described in the following subsections.

Self-Managed Services

One alternative to perform the Logistics and Warehousing services is that the NaaS operator implements them with its own resources and facilities. This option is only applicable if the NaaS operator already has an existing implementation of these services or if it possesses enough resources to implement them. Another case that fits this strategy is for small NaaS Operators with no more than a handful of sites as little space to store equipment is required.

The implementation of this option requires an additional design phase if the Logistics & Warehousing services are not already deployed. Furthermore, it may represent additional costs such as Warehouse rent of space and specialized Management Systems.

More details on the processes related to Logistics & Warehousing design and implementation are included as part of the Logistics & Warehousing Module.

3rd Party Managed Services

Another option for the NaaS operator consists of hiring a 3rd party Logistics and Warehousing Company to manage these services. In this way, the operations of these services are the responsibility of the 3rd party, lightening the operational workload of the NaaS operator.

This option is recommended for the NaaS operator as the burden of these services will decrease when the deployment phase is completed and only will apply for the spare equipment.

Equipment Vendor Services

Some vendors offer Transportation & Logistics services for their equipment, so NaaS operators can rely on them. However, this only applies to certain regions and vendors that include these services as part of its purchase agreements. If the vendor provides these services as an additional feature, the considerations of the 3rd Party Managed Services are applicable.

5.1.5.2 Infrastructure

The Infrastructure strategy describes the model that NaaS operators will implement to build and manage the required network infrastructure (e.g. Fiber Infrastructure, Sites, and Towers). The following subsections examine different models to implement this strategy.

Own Infrastructure

The NaaS operator can perform these activities with its own resources if it already has an existing infrastructure (e.g. MNO with existing mobile network, Tower/Fiber Companies) or if it possesses enough resources (e.g funding and personnel) to deploy them.

The development of this option requires an additional design phase for the new infrastructure to be deployed and additional time to acquire the sites. Moreover, it may represent high additional costs such as material (e.g. fiber optic, steel), machinery and equipment construction.

More details on the processes related to the infrastructure design and implementation are included as part of the Tower/Site construction and Fiber Construction modules.

Infrastructure Sharing

Infrastructure sharing models can have a profound, positive impact on the economics of network deployment into rural and remote areas. In this model, the involved partners share the capital and investment costs of the network infrastructure deployment. For NaaS operators, potential partners to be considered by the NaaS operator are MNOs with no coverage in the area, Tower or Fiber companies.

When the infrastructure is already deployed, the operation and maintenance tasks can be shared or can be absorbed by one of the parties as part of the initial agreements.

Leased Infrastructure

If there is infrastructure available in the area, the NaaS operator can lease the infrastructure to a 3rd party (e.g. MNO with 2G/3G coverage, Tower/Fiber Companies). In this form, the NaaS operator can avoid investment in new infrastructure deployment. The most common type of leased infra is in the form of telecom towers and transport networks.

Additionally, the operation and maintenance of the infrastructure are absorbed by the 3rd party, lightening the operational workload of the NaaS operator.

5.1.6 Operations Strategy

The operations strategy defines the model that NaaS operators will follow to perform the network operation and maintenance tasks. This includes the management of the Network Operation Center (NOC), Field Maintenance and Service Operations Center (SOC), as required. Figure 11 displays the different alternatives related to the Operations Strategy. The NaaS operator must evaluate each alternative and select the most appropriate according to its conditions. The final decision must be recorded in the Strategy & Scope Definition Template.

Figure 11. Operations Strategy Alternatives

Self-Managed Services

The NaaS operator can perform the network operation and maintenance activities with its own resources and facilities.

However, it represents additional costs for the required specialized staff to perform the activities. Furthermore, it may represent additional costs such as rent of space and specialized Network Systems.

3rd Party Managed Services

An alternative option is to hire a 3rd party Company that manages the network operation and maintenance activities. In this way, these activities are the responsibility of the 3rd party, lightening the operational workload of the NaaS operator.

This model is also convenient when sites are deployed in rural locations that are far away the locations the NaaS operates from. Therefore, the disposal of local resources is also a deciding factor to select a 3rd party company.

In this scenario, the cost of specialized hardware cannot be avoided because most of the network equipment requires a Licensed Software to perform the network operation and maintenance activities.

5.1.7 Organization Design

The organization design specifies the division of functional areas within the NaaS operator structure. It establishes the different levels of hierarchy within the organization and delimits the responsibilities of each of them.

Figure 12 presents a typical NaaS operator organizational structure. This structure includes a Leadership Team which develops the Strategy & Scope and the High-level Project Plan.

Figure 12. Typical NaaS operator Organizational Structure.

displays five areas that are required for the successful operations of the NaaS operator:

  • Engineering: Responsible for the generation of the HLD/LLD Network Designs for the different network segments and technologies. It also performs the technical evaluation of different vendor equipment solutions.
  • Deployment: Manages the activities relevant to the network deployment including among others the installation & commissioning and integration of the network equipment.
  • Operations & Maintenance: Oversees the activities related to network operation and maintenance, including among others the monitoring of the network and the incidence response.
  • Commercial / Business Development: Supports with commercial related activities such as the procurement process and vendor management.
  • Financial & Legal: Controls financial resources of the NaaS operator as well as provides the legal support required.

Depending on the size of the NaaS operator, the management and supervision of the available resources can be done by one person for all the areas. However, in larger organizations, a manager for each function might be required. The specific resource requirements for each of the presented areas are examined in more detail in the High Level Project Plan Module.

5.2 Long-Term Strategic Objectives Definition

A long-term strategic objective is a long-term (e.g. 3-5 years), broad, continuous statements that holistically address the different areas of the NaaS operator. Long-term objectives that are viable and appropriate should remain as something that will not change and must be achieved.

The long-term strategic objectives should consider the SWOT analysis. In this way, they will be built on the strengths of the company, fix weaknesses, pursue opportunities and resolve any threats.

In order to define the long-term strategic objectives, the key activities that need to be performed to achieve the vision must be identified considering the strategies presented in section 5.1.

An example of a long-term strategic objective in the NaaS operator context is:

  • To expand network coverage for 10,000 people in the rural La Selva Lacandona area in the following 3 years.

5.3 Set Organization-Wide Goals and Measures

Once the long-term objectives have been identified, they must be translated into goals and measures that can be clearly communicated to the leadership team. Effective goals clearly state what, when, how, and who, and they are specifically measurable. They should address the tasks that need to be performed in the short-term (1-3 years) to achieve the strategic objectives.

The organization-wide goals should present the following characteristics:

  • Specific: Goals need to be specific as they answer the questions of How much? and What kind.
  • Measurable: Goals must be stated in quantifiable terms. Measurable goals facilitate management planning, implementation, and control. For example, a measure might be # of new customers or % complete.
  • Attainable: Goals need to be achievable. The goals must be realistically reached by the NaaS operator
  • Responsible person: Goals must be assigned to a person or a department. The goal responsible needs to be the point person who will ensure the goal is achieved.
  • Time-specific: With reference to time, goals must include a timeline of when they should be accomplished.

Examples of long-term strategic objectives in the NaaS operator context that comply with the SMART characteristics are:

  • Design team must deliver the RAN and Transport Designs for 20 sites in the following year.
  • Deployment team must deliver 20 operational sites in the following year.

6 Organizational Governance & Performance

This section includes methodologies to control and manage the implementation of the defined strategies. The governance definitions aim to communicate the overall strategies as well as measure and monitor progress towards strategic targets.

6.1 KPI Definition

Key Performance Indicators (KPIs) are the critical indicators of progress towards an intended result. KPIs provide a focus for strategic improvement and create an analytical basis for decision making.

In order to generate a KPI for the strategy implementation, the following elements must be defined:

  • Units of measure: Identify the units of measures associated with each goal. If this cannot be performed, a standard measure should be picked to track the goal.
  • Target: The target is the numeric value that needs to be achieved. This target must be set according to the units of measure and the timeframe in which need to be accomplished.
  • Source: Clearly identify where the source of the reporting measure and who is responsible for reporting on it.
  • Frequency: Is the frequency of the reporting measurement for the KPI. The recommended frequency is to review the strategy performance on a monthly basis.

A useful form to present the KPIs is through the use of the Strategy Scorecard as displayed in Table 4. The NaaS operator should construct a similar Strategy Scorecard with the specific objectives KPIs for each of the Strategies presented in Section 5.1.

Strategy Name

Goal

Units of Measure

Target

Frequency

Source

Goal 1

Goal 2

Table 4. Strategy Scorecard Format

The aim of the Strategy Scorecard is to identify adjustments over the time according to the KPIs behavior.

An example of the definition of a KPI in the NaaS operator context is presented in Table 5. KPI Example.

Deployment Strategy

Goal

Units of Measure

Target

Frequency

Source

Operational Sites

Number of sites

9 sites

Monthly

Project Plan

Subscribers

Number of subscribers

3300 subscribers

Annually

Project Plan

Data Volume

Aggregated Mbps

600 Mbps

Monthly

Project Plan

Table 5. KPIs Example

6.2 Establish Schedule for Progress Reviews

Holding regular strategy reviews is the key to the implementation of the Strategies. These meetings will give the ability to manage activities that drive future results and hold people accountable for making sure those activities happen. Furthermore, the adopted strategies can change based on competitors and government actions, therefore, strategy meetings are important as the world around the NaaS evolves.

A Schedule for Progress Reviews must be defined, describing the specific dates and communication channels to communicate with stakeholders regarding the strategic plan and the associated progress. This schedule should be employed by strategy managers in the organization to ensure the awareness about the intent and progress of the strategic plan is fully understood by key stakeholders.

The following steps must be perform to define the Schedule for Progress Reviews:

  • Set up monthly strategy meetings. Monthly strategy meetings dont need to take a lot of time but it is important that key team members report on their progress toward the goals they are responsible for, including reporting on metrics in the scorecard they have been assigned. Furthermore, a frequency greater than quarterly to review the strategy performance must be avoided.
  • Finally, set up annual strategic review dates including new assessments and a large group meeting for an annual plan review.

6.3 Goal Tracking Methodologies

The NaaS operator must establish the performance management review system in order to track the progress towards the goals to identify possible issues and make course corrections if necessary.

One alternative is to use the Strategy Scorecard method presented in Section 6.1 as a performance management review system. In this way, the Strategy Scorecard must be presented by the responsible people during each meeting.

Additional tools such as management and technology systems can help track the progress of the plan and make it faster to adapt to changes. As part of the system, milestones can be built into the plan that must be achieved within a specific time frame.

6.4 Review and Adapt Process

A Strategic Planning Retreat must be held at least once per year. During this session, all the elements in the Strategy, including the Mission and Vision Statements must be reviewed in order to confirm that are current and still relevant to the NaaS operator.

The following elements must be considered to perform the Strategic Planning Retreat:

  • Establish a Specific Deliverable: Create a very specific description of exactly what the deliverable of the event looks like.
  • Name a Facilitator: The facilitator ensures people stay on time and speak about topics that are relevant and on task. The NaaS operator must find a facilitator who understands and runs strategic planning meetings regularly.
  • Research: A research about customers needs and competitors actions must be conducted prior to the retreat. This will support the decision-making process.

After updating the Strategy Plan, it must be distributed to all elements within the organization to align the priorities and ensure that everyone knows what they are responsible for.


1 Introduction

Development of the High-Level Project Plan is the immediate next step after Strategy & Scope definition of the NaaS Initiative. Together with the NaaS Strategy, the High-Level Project Plan serves as a roadmap to stand-up the NaaS Operator. It provides the reference plan to manage and control critical tasks. It also provides task and timeline guidance for detailed planning associated with Design, Deployment and Operation of the Network.

This module describes the key planning areas and components. It aims to transform the Strategic Plan into an actionable plan in the key areas of Schedule, Resources and Communication Plans.

Together, these elements comprise the High-Level Project Plan, which is the catalyst for action and provides the NaaS Operator and key stakeholders a holistic view of what needs to be done, the resources required, and a guidance on how to communicate the plan across the NaaS Organization.

1.1 Module Objectives

This module will enable NaaS Operators to establish a High-Level Project Plan to organize and coordinate all the activities required to build and turn-up a NaaS Initiative. The specific objectives of this module are to

  1. Provide an information base of fundamental planning aspects for a NaaS Initiative, creating awareness of the interactions among these aspects and NaaS Strategy.
  2. Guide the NaaS Operator to identify all the required resources and generate an efficient plan for resource acquisition and utilization.
  3. Provide guidelines to identify key tasks and milestones, and generate a High-Level Schedule that aims to achieve NaaS Operators strategic goals and objectives.
  4. Support NaaS Operators to define a Communications Plan that provides an Organization-wide structure for meetings, reporting, communication channels and documentation.

1.2 Module Framework

The Module Framework in Figure 1 describes the structure, interactions and dependencies among different NaaS operator areas.

Strategic Plan & Scope is the first stream to build the NaaS Operator. The modules in this stream aim to drive strategic decisions that impact all the downstream modules. From this point, High-Level Network Architecture can be defined, establishing the guidelines for Network Design which in turn kick-starts the deployment phase.

High-Level Project Plan comes after Strategy & Scope definition. However, there is an interaction between the outputs of these modules, in the sense that High-Level Project Plan may trigger updates to Strategy & Scope.

Figure 1. Module Framework

Figure 2 presents the High-Level Project Plan functional view, where the main functional components are exhibited. Each component is discussed in its own section of the outline, detailing the impact on other blocks and a process to generate an integrated High-Level Project Plan.

Figure 2. High-Level Project Plan Functional View

The rest of this module is structured as follows: Section 2 presents an overview of the NaaS Planning components and its interaction with the Strategy. Sections 3 to 5 provide deeper analysis and guidance to perform actual Resource, Schedule, and Communications Plan in the context of rural NaaS initiatives.

2 High-Level Project Plan Overview

This section presents the key aspects and components to elaborate a High-Level Project Plan, emphasizing their relevance for the NaaS Operator and interdependency with the NaaS Operator Strategy & Scope.

2.1 Strategy & High-Level Project Plan Interaction

Strategy and scope of the NaaS Operator are first established to set the route that the NaaS Initiative will take and the general methods to be used to reach the long-term objectives.

However, the only way the Strategy gets executed is to align resources and actions through an effective Project Plan that translates the Organization Strategies and Objectives into structured tasks and timelines that fit available resources.

Outputs from Strategic Planning are Organization Strategies and Long-Term Strategic Objectives (See Strategy & Scope Module). These outputs are the starting point and provide a general guide for the High-Level Project Plan, which aims to answer how the Strategy and Objectives will be achieved and who must do what to accomplish them.

Based on the above, it is fair to say that the Strategy drives the Project Plan. However, in todays rapidly changing environment, NaaS Operators may find during the development of the Project Plan that certain assumptions are not valid or that additional constraints that werent considered during Strategy definition must be incorporated. Thus, an iterative planning approach must be applied in which findings from the High-Level Project Plan triggers updates to the Strategy restarting the planning cycle until the NaaS Operator approves both Strategy and Project Plan.

This module enables NaaS Operators to establish a High-Level Project Plan to organize and coordinate all the activities required to build and turn-up a NaaS Initiative based on the NaaS Operator strategy.

2.2 NaaS Planning Fundamentals

In the following sections, an initial description of planning components is provided to build a holistic view of the planning process.

2.2.1 Resource Plan Overview

NaaS Operators require different types of resources to plan, deploy and operate the network and the services it provides. These resources can be categorized as follows:

  • Spectrum. RF spectrum required for the LTE Radio Access Network and for microwave links. This resource is always required and is critical for any NaaS Initiative. These resources must be secured early in the Project to eliminate the risk of delays due to spectrum issues.
  • Telecom Infrastructure. Includes network sites (RAN, aggregation, core), fiber, cabling and grid power infrastructure. In particular, RAN sites are of major importance as they are the basis for network deployment. Depending on the strategy and available infrastructure, NaaS Operators may reuse existing infrastructure or build new (Greenfield) sites. In any case, this should be considered a high priority item and the Project Plan must accommodate reasonable timelines for site acquisition and permitting.
  • Network Equipment. This term refers to the equipment that ultimately comprises the network and supports its services, including RAN, transport, core, and power equipment. Network Equipment is usually acquired through an RFI/RFP process after the network designs have been accepted.
  • Installation & Construction Materials. Physical resources required for site construction or adaptation, fiber construction, and network equipment installation. Construction materials are required only for Greenfield sites, fiber infrastructure construction and whenever the NaaS Operator is adapting existing sites. On the other hand, installation materials are always required. Depending on the NaaS Operator Strategy and agreements with contractors, these materials could be provided by Construction and Deployment contractors, respectively.
  • Tools. This term includes construction tools, installation and commissioning tools and other types of tools such as vehicles. Specific tools required vary for each initiative based on NaaS Operator Strategy, mainly in accordance with the activities that the NaaS Operator is supposed to perform with its own staff. These tools are often provided by the 3rd Parties in charge of construction, site surveys, and installation and commissioning.
  • Systems. NaaS initiatives require engineering and network operation systems in addition to typical Business/Enterprise systems for the overall management of the organization. Systems can be sourced in various ways and low-cost or opensource alternatives can be considered.
  • Facilities. In addition to network sites, there might be other facilities required such as headquarters, warehousing facilities, and network operation center (NOC). Specific facilities required depends on the NaaS Operator strategy and size of the network. These should be acquired at the right time through buy or leasing agreements.
  • Staffing. This category includes all the human resources that are part of the NaaS Operator Organization. Specific departments and positions are part of the Strategy and may include the Executive Team along with Deployment, Engineering, Operations, Commercial, Financial and Legal Teams. Adequate staffing plan must consider appropriate timeframes for resource onboarding as well as identification of temporary vs long-term resources. Staffing is the most important resource, as the personnel of the NaaS Operator is the foundation for a successful deployment and operation; without a good team, the best plan will fail.

In addition to the resources above, services are considered as part of the Resource Plan. However, it is important to note that services are not the same as resources. Services are provided by 3rd Parties and are used to satisfy a requirement over a period of time, i.e. they are temporary. In some cases, instead of acquiring a specific resource NaaS Operators may decide to contract a service that replaces the required resource.

  • Services. Contemplates all the 3rd Party services that according to the Strategy will be required by the NaaS Operator. Potential services to be considered are connectivity, deployment, construction, power, warehousing, and logistics services. Contracting of services may follow an RFI/RFP process and formal agreements must be in place specifying all the related terms and conditions.

The Resource Plan specifies how the above resources should be categorized, acquired, managed, and released. The Resource Plan for NaaS Operators must include at minimum the following items:

  1. Resource Identification. Identification and quantification of the required resources to accomplish objectives.
  2. Resource Acquisition. Plan and guidelines on how the resources will be acquired. A separate section should be considered for staffing in which specific positions and responsibilities should be taken into account.
  3. Resource Management. Details on how and when resources will be utilized, considering guidelines for allocation, control, and release of resources.

Further details and guidance for the development of the Resource Plan are provided in Section 3.

2.2.2 Schedule Overview

Project scheduling provides a plan that describes when the Project Milestones will be delivered and the activities involved to achieve these milestones, including duration and dependencies among activities. As part of the High-Level Project Plan, the Schedule serves as a tool for communication, managing stakeholders expectations, and as a basis for reporting.

As can be expected, there is a myriad of tasks that must be planned and managed throughout the Lifecycle of the NaaS Initiative. This Module contemplates activities from NaaS Initiative launch up to the completion of Deployment and entrance into Operations Phase. At a high level, the milestones and tasks can be grouped into three categories:

  • Resource Acquisition and Organization Setup. Includes all tasks and activities required to complete the project plan, set-up headquarters, concrete key agreements, acquire spectrum and complete staff onboarding.
  • Network Planning & Design. Comprises tasks and activities that aim to define the network architecture, develop designs and select vendors, culminating with the purchase order for network equipment and the low level network design.
  • Network Deployment & Start of Operations. This category is tied to site deployment activities from site acquisition, construction, hardware delivery, installation and commissioning that ultimately contribute to put sites-on air and initiate network operations.

Specifying the Project Schedule is of paramount importance to manage the timely completion of the Project and the achievement of the NaaS Initiatives goals and objectives. Through the analysis of activities sequences, durations, resource requirements, and schedule constraints the project schedule creates a baseline for project execution, monitoring and controlling.

Conditions of every NaaS Operator are different; thus, specific tasks will vary across the NaaS spectrum based on Operators Strategy, existing resources, size of the network, financial constraints, and technology requirements. In a similar vein, task sequencing, interdependencies and duration may vary as well.

A comprehensive assessment of these variations and a method to develop a Project Schedule is presented in Section 4 through the analysis of a generic schedule. In addition, instruction for customization of the generic schedule is provided, including task duration estimation, task sequencing and schedule validation.

2.2.3 Financial Analysis Overview

A critical component of the High-Level Project Plan is a Financial Analysis or Business Model that models the financial performance of the Initiative. It is the primary tool to get investors to buy-in the Project and obtain private or public funding. Ultimately the Financial Analysis will be the baseline for comparison and the decision instrument to approve the Project Plan.

Based on NaaS Strategy and Objectives, and other components of the Project Plan, the financial analysis presents forecasts for costs, revenues, profits and cash flows, and obtains financial metrics such as return-on-investment (ROI), Break-even Year, net present value (NPV) and internal rate of return (IRR).

This analysis is usually done during the early phases of the NaaS Initiative, sometimes before the Strategy definition, and refined during the Project Plan. Furthermore, each Organization may have a different method, scope, requirements and set of metrics for the Financial Analysis. In consequence, a detailed treatment of the Financial Analysis is out of the scope of the PlayBook and it is left to the NaaS Operator to use the methodology and metrics that best fit their needs and requirements.

In any case, it is important to consider at least the following inputs for the Financial Analysis:

  • Network Infrastructure & Equipment CapEx
  • Network Infrastructure & Equipment OpEx
  • Other Capital Expenditure
  • Other Operational Expenditure
  • Labor Costs (Staffing)
  • Sales, General and Administrative Expenses
  • Subscriber & Revenue Forecasts
  • Taxes
  • Hurdle Rates
  • Asset Depreciation

2.2.4 Communications Plan Overview

The Communications Plan is the component of the High-Level Project Plan that documents how communications across the NaaS Organization are structured, implemented and monitored. It defines the mechanisms for communications activities based on information needs, the methods to distribute information and coordinate tasks, and the methods for storage and retrieval of information.

The implementation of an effective Communications Plan provides the following benefits:

  • Maintains coordination of efforts and keeps focus on the things that matter the most across the Organization.
  • Ensures integrity and availability of the information required for the execution, control and monitoring of activities
  • Enables efficient and effective reporting throughout the NaaS Initiative Lifecycle.

For NaaS Operators, these benefits can be translated into one key success factor: it avoids delays and re-work due to miscommunication and lack of coordination. This is of utmost importance to keep the initiative alive and financially healthy due to the high cost of re-work and financial constraints that are faced by rural NaaS Operators.

The Communications Plan for NaaS Operators must include at minimum the following items:

  • Meeting plans and reporting mechanisms, identifying stakeholder information requirements, information to be reported, format and level of detail, timeframe and frequency for meetings and reports.
  • Communication channels consisting of methods or technologies used to convey the information and coordinate tasks, such as phone, email and collaboration software tools.
  • Systems and techniques such as Wikis or cloud Document Management Systems to manage documentation, including key inputs and outputs from organizational processes.

In addition, communication to external entities should be considered, establishing rules and policies to share marketing, technical and other confidential information. External communication should consider any constraints derived from specific legislation or regulation and it is considered out of the scope of this module.

Section 5 presents a detailed view of the Communications Plan providing guidance for the NaaS Operator to build their own Communications Plan.

3 Resource Plan

This section explores the tasks and factors involved in Resource Planning, including guidelines for the NaaS Operator to develop their own Resource Plan.

3.1 Resource Identification & Assessment

The first step towards a Resource Plan is to identify the required resources according to the Scope and Strategy of the NaaS Initiative. This is usually done through brainstorming and resource breakdown techniques using objectives and categories of resources to go from general categories to specific individual resources.

Then, after identifying individual items, each item should be quantified based on the scope and strategic objectives.

To simplify this process, a generic list with the most common resources categories and subcategories, and guidance to identify applicability for the initiative is provided in Table 1. NaaS Operators are encouraged to use this table to identify their required resources.

Category

Subcategory

Description

Applicability Criteria

Spectrum

RAN

Spectrum used by the LTE System. It could be licensed or unlicensed.

Always required

Microwave

Spectrum used by microwave links. It could be licensed or unlicensed

For NaaS Operators using Microwave backhaul

Telecom Infrastructure

RAN Sites

Physical Site and civil infrastructure for RAN equipment deployment. Examples are Towers, Poles, Rooftops. Land plots may be considered as well for Greenfield sites

This category includes cabling and supporting infra to route cables and host equipment such as cabinets, racks, ODFs and trays.

Always required. (Type of site and specific requirements vary by NaaS Operator)

Transport Nodes

Aggregation Sites with existing transport/fiber connectivity that may as well be considered for RAN Installation

Required except for the case of a network using satellite backhaul only with a 3rd Party or Cloud Mobile Core

Datacenters

IT or Telco Facility that hosts Compute, Networking and Storage Infrastructure that supports Systems and virtualized elements such as virtual EPC

Required for private cloud implementation to support a virtualized core and other Systems. Does not apply for public cloud services or basic server deployment at an IT room in the Headquarters or other network sites

Fiber Infra

Civil and Optical outside plant infrastructure used to provide transport connectivity. Includes Ducts, Poles, Fiber Cables, Splicing Boxes, etc.

Required if fiber backhaul is being deployed and/or managed by the NaaS Operator. If using a 3rd Party fiber backhaul service, the NaaS Operator is not responsible for this resource

Power Infra

Power cabling, electrical boards, switches, regulators and grounding

Always required. Might be included as part of the tower/site leasing service

Network Equipment

RAN

LTE Base Stations

Always required

Transport

Last Mile and Aggregation Transport Equipment consisting of microwave radios, routers, switches, and satellite modems (VSAT)

Always required. Equipment varies depending on the transport technology

Power Equipment & Batteries

Components of the systems to power equipment at network sites, including solar panels, controllers, rectifiers, power over Ethernet (PoE) injectors, Batteries, etc.

Always required

Antennas

Various type of antennas required to transmit and receive radio signals. Includes RAN, microwave and GPS antennas

Always required (at least RAN Antennas)

Mobile Core

EPC and Policy and Charging Control nodes, that are usually implemented as virtualized appliances

Depends on Strategy and Mobile Core Architecture. Only applies when implementing a mobile core instead of using a 3rd Party or cloud core

IT infrastructure

Servers and networking components that support systems, which might include virtualized network functions

Required if implementing virtualized elements (e.g. core) or systems in a private cloud environment

Installation & Construction Materials

Cables & Connectors

Cables, connectors and associated materials (e.g. tape) that are used for equipment installation and connection: Electrical, Coaxial (RF) and Optical Cables

Always required. Might be provided as part of a 3rd Party Installation & Commissioning service

Mounting Hardware

Miscellaneous elements required to install and mount network equipment

Always required

Construction Materials

Structural and building materials required for site and fiber construction, including bricks, concrete, blocks, steel, ducts, pipes, etc.

Required for site and fiber construction. Usually provided as part of the service with a Construction Company

Tools

I&C Tools

Tools used throughout the installation and commissioning process, up to acceptance testing such as Smart Phones (UEs) & Test Applications, Power & Hand Tools, RF & Network Testing Tools, Safety Equipment, Communications Equipment

Always required. Usually provided as part of a 3rd Party Installation & Commissioning service

Construction Tools & Machines

Heavy Equipment and other construction tools required for site and fiber construction

Required for site and fiber construction. Usually provided as part of the service with a Construction Company

Vehicles

Cars and trucks for equipment transportation, site surveys, maintenance and any other field task

Required if the NaaS Operator is expecting to perform site activities through its own personnel

Systems

Network Operations

Management, Monitoring, Inventory, Ticketing, Billing

Always Required

Engineering Tools

RF and Microwave Planning Tools for High and Low Level Network Designs

Depends on network size and overall strategy.

Business/ Enterprise Systems

Billing and Administrative Software, Document Management, Project Management, etc.

Systems to be implemented vary by NaaS Operator

Facilities

Buildings

Facilities apart from network nodes such as Headquarters, Warehousing, and network operation center (NOC)

Headquarters are always required. Other facilities can be avoided by using 3rd Party Services or converged at Headquarters

Office Equipment & Furniture

Required items at the office: Desks, chairs, PCs, Printers, Phone, etc.

Always required. Specific items will depend on NaaS Organization size, structure, and scope

Staffing

Executive Team

Directors and/or managers responsible for the overall management (Managing Director) and specific departments such as Technical, Operations and Commercial

Always required. Size and specific functions may vary

Engineering Team

Personnel in charge of network planning and engineering

Always required. Size and scope may vary

Deployment Team

Personnel in charge of coordinating and executing site deployment, installation, commissioning and integration

Always required. Size and scope may vary

Operations Team

Personnel in charge of network monitoring, maintenance, fault management and ensuring service availability and performance

Size and scope may vary and sometimes can be merged with Deployment Team.

Commercial / Sales

Personnel in charge of engaging with customer MNOs and other business relationships

Not always required, depending on size and Organization Strategy: Functions may be converged with Financial/Legal

Financial / Legal

Personnel in charge of procurement and legal aspects in the Organization

Not always required, depending on size and Organization Strategy: Functions may be converged with Commercial/Sales. Legal functions might be outsourced

Others (HR, Compliance, etc.)

Personnel in charge of other supporting and ancillary functions such as human resources, performance, compliance

Depends on Organization Strategy

Services

Transport & Internet Connectivity

Enterprise and network connectivity services acquired from a 3rd Party. This includes Terrestrial and Satellite backhaul as well as Internet Connectivity

Always required

Deployment

Equipment transportation, Installation and commissioning services for site deployment

Depends on Organization Strategy

Construction

Civil works to build or adapt a network site or to deploy fiber outside plant infrastructure

Depends on Organization Expertise and Scope

Grid Power

Power service from an electrical company

Always required

NOC

Managed service for network Operation that may include different tiers and functions such as monitoring, ticketing, service desk, fault management and incident management

Depends on Organization Strategy and size of the network

Warehousing & Logistics

Hardware transportation and warehousing service

Depends on warehousing requirements, geographical span of the network and NaaS Organization Strategy

Table 1. Generic List of NaaS Resources.

In addition, to detail staffing resources, Table 2 provides generic positions and key activities that can be easily adapted to the specific NaaS Organization.

Position

Reports To

Key Activities

Managing Director

Board / Investors

.Overall Management
.Strategy Definition & Implementation

Technical Director

Managing Director

.Lead network and systems decisions
.Oversee Engineering, Deployment and Operations

Financial/ Commercial Director

Managing Director

.Manage legal, commercial and financial activities

Engineering

Engineering Manager

Technical Director

.Oversee and engage in network planning, design and optimization activities

RAN Engineer

Engineering Manager

.RAN site planning, design and optimization

Transport Engineer

Engineering Manager

.Transport network planning, design and optimization

Core Engineer

Engineering Manager

.Mobile core planning, design and optimization

Deployment

Deployment Manager

Technical Director

.Planning, control and monitoring of network deployment activities

Deployment Engineer

Deployment Manager

.Inventory, warehousing and kitting
.Site acquisition and construction overseeing
.Coordination of I&C tasks

Integration Engineer

Deployment Manager

.Support field activities
.Remote site integration activities

Deployment Technician

Deployment Engineer

.On-site Installation & Commissioning

Operations & Maintenance

O&M Manager

Technical Director

.Lead network operations, preventive and corrective maintenance
.Manage implementation of network and IT Systems

Systems/IT Engineer

O&M Manager

.Implement, operate and maintain network and IT Systems

NOC Operator

O&M Manager

.Monitor network status and performance
.Fault Management and Troubleshooting
.Coordination of field activities

Site Technician

NOC Operator

.Preventive and corrective field maintenance

Commercial / Business Development

Sales Manager

Financial/ Commercial Director

.Manage engagement with customers

Sales Executive/ Representative

Sales Manager

.Point of contact for customers
.Billing

Procurement Manager/Analyst

Financial/ Commercial Director

.Supply chain management

Financial & Legal

Financial Specialist

Financial/ Commercial Director

.Accounting and payments

Legal Associate

Financial/ Commercial Director

.Manage agreements, permits and other legal issues
.Oversee regulatory aspects

Other

Administrative Support

(May vary)

.Human Resources
.Report preparation
.Paperwork

Table 2. Generic Staffing Requirements

Once the required resources have been identified and quantified, NaaS Operators must assess their existing resources and perform a gap analysis to flag those resources that must be included in the Resource Acquisition Plan.

Its worth noting that in the Telecom industry its common to use 3rd Party Resources for temporary works or specific tasks requiring certain expertise, specially when these resources are required for a small amount of time. This would lie under the services category.

Existing resources are assessed by applying the following questions:

  • Is the quantity of existing resources enough to cover the NaaS requirements?
  • Are these resources available or can they be re-purposed for this initiative?
  • Is the quality of existing resources adequate to comply with the requirements?
  • What actions should be taken to utilize existing resources for the NaaS Initiative?

Based on the assessment, the delta with respect to the required resources should be calculated and the type and amount of resources that must be acquired registered.

To support NaaS Operators throughout Resource identification and assessment, a Resource Checklist Template and a Staffing Plan Template are provided for customization and fill-out by the NaaS Operator.

3.2 Resource Acquisition Plan

After getting a clear view of the required resources, NaaS Operators must develop a plan to acquire them, which should consider not only financial constraints but also time and quality constraints as well as NaaS Operator Strategies and Priorities. In the case of Staffing, resources need to be trained and this training must be considered in the plan.

Sections below examine possible acquisition approaches for each resource category, providing general guidelines for the NaaS Operator to establish their acquisition approach.

3.2.1 Spectrum Acquisition

Spectrum may be acquired by re-farming of spectrum owned by the NaaS Operator, licensing through standard or rural-specific mechanisms / grants; leasing or sharing agreements with MNOs or simply using unlicensed spectrum.

The selection of the acquisition approach is mainly determined by the Strategy. Further guidance is provided in the Strategy & Scope Module.

The spectrum must be secured early in the project as this is an absolute requirement and impacts the overall design of the network. If the primary option for acquisition fails, an alternative approach can be applied.

3.2.2 Telecom Infrastructure

If the NaaS Operator does not own the required telecom infrastructure, the most time-efficient approach for acquisition is through lease agreements with infrastructure owners, in particular for network sites such as RAN, aggregation and core sites. Often, a single agreement may include several infrastructure assets; for example, when leasing space in a tower, this may include power, cabling, and other supporting infrastructures such as shelter and rack space.

A specific case of infrastructure leasing is optical infrastructure sharing where only parts of the required infrastructure are provided through the lease agreement (e.g. poles) and the rest (fiber cable, splicing boxes, etc.) is deployed/provided by the NaaS Operator.

On the other side, infrastructure can be built, which further presents the option to do it through a 3rd party service. Building infrastructure requires a significant amount of effort and upfront investment, carrying a high associated risk in case the initiative fails to meet business targets.

Thus, NaaS Operators should privilege the reuse of existing infrastructure through leasing or sharing agreements. Building infrastructure should only be considered when no infrastructure is available for lease or for strategic reasons such as avoiding specific contractors or high lease prices.

3.2.3 Network Equipment

Network equipment must be acquired from an Original Equipment Manufacturer or a Distributor through a formal Procurement Process that involves the issuance of Requests for Information (RFI) and Requests for Proposal (RFP) that allows to compare solutions from various vendors at the commercial and technical level.

This is the recommended option as this ensures a healthy competitive environment and that the selected equipment meets technical and commercial requirements from the NaaS Operator. However, in financially challenged environments, procurement of refurbished equipment from MNOs migrating technology in urban areas can also be considered.

3.2.4 Installation & Construction (I&C) Materials

The simplest way to acquire installation and construction materials is to include them as part of the agreements for I&C or construction services. However, based on the strategy for these agreements or when the NaaS Operator strategy is to take over I&C and/or construction activities, these materials shall be procured by the NaaS Operator.

To buy materials for I&C and construction, the NaaS Operator can use a local, regional or global distributor. In this case, a full-fledged RFI/RFP process may not be required depending on the type of material, the volume and whether the NaaS Operator has previously onboarded a suitable vendor.

Finally, some materials like mounting hardware could be in-house built if the NaaS Operator has the necessary tools and expertise. The recommendation for NaaS Operators is to include these materials as part of the agreements for I&C and construction services whenever possible. Otherwise, compare prices from local and regional distributors and assess the need for an RFI/RFP Process.

3.2.5 Tools

Different approaches might be taken based on the subcategory of tools. For I&C tools, a similar approach to that in section 3.2.4 must be considered.

Construction Tools & Machines may also follow the approach in section 3.2.4 with one addition: these tools and machines may also be leased from a 3rd Party. If the NaaS Operator is taking over construction, the most cost-efficient approach should be selected.

When vehicles are required, the NaaS Operator can either buy them or contract a fleet leasing service. If fleet leasing is available in the area, a cost-benefit analysis should be performed to select the best option.

3.2.6 Systems

Systems can be acquired through freeware/opensource software or proprietary software. In addition, they can be implemented through cloud services or deployed on-premise.

As a rule of thumb, NaaS Operators with small networks and tight financial constraints may decide for on-premise opensource systems. If that is not the case, the Operator should prefer cloud solutions due to setup, maintenance and cost efficiency. Nevertheless, other factors may influence the decision and should be taken into account by the NaaS Operator, such as regulatory constraints, cloud connectivity issues and the pre-existence of on-premise solutions and skills to set up and maintain them within the Organization.

3.2.7 Facilities

The first option for facilities acquisition is to re-purpose or reuse facilities previously owned by the NaaS Operator. If these do not exist or they cannot be reused, then the leasing of facilities can be considered. In the latter case, the NaaS Operator should look for the best conditions and try to find facilities with the required furniture and amenities. Managed facilities should be preferred.

The last option would be to buy/build facilities which is not recommended due to the lengthy process and associated financial risk.

3.2.8 Staffing

Human resources have a key role in the overall performance of the Organization and the achievement of the objectives, thus, a recruiting process should be put in place considering a thorough candidate evaluation that includes qualifications, expertise, attitude, and hard and soft skills. In addition, a training plan must be considered to homologate skills and knowledge across teams.

Candidate sourcing can be done through direct references or job posting / search in digital platforms such as LinkedIn, Glassdoor or Indeed or even through social networks (e.g. Facebook, Twitter). In addition, recruiting services may also be considered to outsource candidate search, shortlisting and screening.

The recommendation for NaaS operators is to go with job posting/search in digital platforms and direct references. Recruiting services should be considered when a big organization structure has been defined along with aggressive deployment plans.

For temporary roles, NaaS Operators should also evaluate the option to have contractors or temporal workers with the required expertise to perform specific tasks such as low level engineering, integration and field deployment (Installation & Commissioning).

3.2.9 Services

In general, 3rd Party services must be acquired from contractors and service providers through a formal RFI/RFP process that allows to compare solutions from various vendors at the commercial and technical level.

This is the recommended option as this ensures a healthy competitive environment and that the selected contractor meets technical and commercial requirements from the NaaS Operator.

The selected acquisition approach for each resource will drive the tasks and time required to secure resources. For example, when leasing space in RAN sites, the process is faster than building new sites. Leasing only requires performing a site survey, negotiating terms, and signing agreements which could be a 4 to 8 weeks process. On the other hand, building a new site requires land acquisition, site permitting and site construction that together can take from 3 to 9 months.

NaaS Operators must estimate acquisition tasks and timelines, establishing deadlines that will be taken as inputs for the Schedule. It is important to note that timelines for these tasks may vary by region and the NaaS Operator must estimate them based on past experience and inputs from the environmental scan performed as part of the Strategy definition (See Strategy & Scope Module).

The provided Resource Plan Template can be used by NaaS Operator to document this analysis

3.3 Resource Utilization Plan

Once acquired, resources must be managed, i.e. they are allocated, controlled and released for the different activities across the NaaS Organization. In addition, some resources are required permanently or for a long term while other resources are sporadically required or required for a specific period only.

Thus, the NaaS Operator shall develop a utilization plan through the following steps:

  1. Develop utilization estimates. For each resource, estimate the timelines in which it will be required.
  2. Assign resources to tasks. Identify the tasks that make use of each resource.
  3. Optimize resources. For certain types of resources, it might be possible to reduce the quantity if the timelines are extended. In other cases, temporary resources can be re-purposed to improve utilization and avoid the overhead of acquiring and releasing too many resources. For example, personnel executing deployment functions can be trained and transferred to Operations once deployment has been completed.
  4. Designate the area or role that will manage each resource.

After performing the analyses in Sections 3.1, 3.2 and 3.3, NaaS Operators can make use of the provided Resource Plan Template to register their Plan.

4 High-Level Schedule

This section provides an examination of key tasks and milestones for NaaS Projects, providing guidance to generate and validate the High-Level Schedule for the Initiative.

4.1 Milestone Definition & Task Identification

Section 2.2.2 stated that there is a myriad of tasks that must be planned and managed throughout the Lifecycle of the NaaS Initiative to accomplish strategic objectives.

Thus, the initial process to establish a Schedule is to identify and document the specific actions to be performed to achieve these objectives.

First, the Project milestones are defined. A milestone is a reference point or major event that marks an important achievement related to the objectives and serves to report progress. These milestones can be derived by identifying the major components of the deliverables or objectives.

Then, high level tasks are identified based on objectives and milestones. This can be done by the NaaS Operator through the decomposition technique that consists in dividing and subdividing milestones and deliverables into smaller parts and actions until getting to the required level of detail.

For the High-Level Schedule is not necessary to get deep into the details and is enough to identify tasks at the Organization Level. However, for further refinement of the Plan, the same technique can be applied to cascade tasks up to the individual level.

Table 3 below provides a generic list with the most common milestones and tasks by category, including guidance to identify applicability for the Initiative. NaaS Operators are encouraged to use this table to define and orchestrate milestones and tasks for their Initiatives.

Category

Task/Milestone

Applicability Criteria

Milestones

Resource Acquisition and Organization Setup

NaaS Initiative Launch

Applies for all initiatives

Project Plan Ready

Applies for all initiatives

HQ Set-up

Required for new companies with no headquarters or for expansions

TX Provider Agreements

Required when relying on 3rd Parties for transport services

Infra Sharing Agreements

Required when tower or fiber infrastructure is going to be leased/shared with a 3rd Party

1st Customer Agreement

Applies for all initiatives

Minimal Team Onboarding

Applies for new organizations

Spectrum Acquired

Applies for all initiatives

Full Team Onboarded

Applies for all initiatives

Network Planning & Design

Network Architecture Defined

Applies for all initiatives

HLD Ready

Applies for all initiatives

Construction Company Onboarded

Required only when greenfield sites are being built through a contractor

Deployment Contractors Onboarded

Required only when site I&C is performed through a deployment contractor

Equipment Vendors Onboarded

Applies for all initiatives

Equipment PO Issued

Applies for all initiatives

Network Deployment & Start of Operations

Sites Acquired

Applies for all initiatives

Deployment Kickoff

Applies for all initiatives

Sites Built

Required only if greenfield sites are being built

1st Site On-Air

Applies for all initiatives

Deployment Complete

Applies for all initiatives

Tasks

Resource Acquisition and Organization Setup

Build Project Plan

Applies for all initiatives

Commercial Agreements

Applies for all initiatives. Required agreements may vary

Spectrum Acquisition

Applies for all initiatives

Headquarters Set-up

Required for new companies with no headquarters or for expansions

Hiring & Staffing

Required for new companies or if additional staff is required

Other Resource Acquisition

Applies for all initiatives. Required resources may vary

Network Planning & Design

Network Architecture Definition

Applies for all initiatives

Network HLD

Applies for all initiatives

Vendor Selection & Onboarding

Applies for all initiatives. Required vendors may vary. At least Network Equipment Vendor is required

Site Acquisition & Permits

Applies for all initiatives

Network LLD

Applies for all initiatives

Hardware Delivery

Applies for all initiatives

Network Deployment & Start of Operations

Training & Preparation

Applies for all initiatives

Site & Fiber Construction

Only for initiatives that require to build greenfield sites or fiber deployments

Site Deployment

Applies for all initiatives

Table 3. Generic NaaS Milestones & Tasks.

4.2 Schedule Generation

Once the Project tasks and milestones have been identified, specific task duration and sequencing must be assigned to orchestrate the High-Level Project Schedule.

There are several software tools in the market, licensed and opensource, to aid and assist operators in this task. Examples of these tools are Microsoft Project or Project Libre, where users can build a new project, add or remove activities, adjust their relationships, durations, and allocated resources. Other available tools are listed in the Runbook Tool Catalogue.

The following sections provide generic instructions to perform these processes. Afterward, a generic view of the High-Level Schedule for NaaS Operator is presented along with guidelines for customization.

4.2.1 Task Duration Estimation

In this process, the NaaS Operator estimates the amount of time each activity will take to complete with the estimated resources.

Estimating activity durations uses information from the Strategy & Scope module, estimated resource quantities, and resource utilization plans. Other factors that may influence the duration estimates include constraints imposed by the strategic goals and objectives.

It’s important to note that the number of resources that are expected to be available to accomplish an activity, along with the skill proficiency of those resources, may determine the activitys duration.

Common techniques to estimate duration are:

  1. Analogous Estimating: Estimate duration or cost using historical data from a similar activity or project. It is a gross value estimating approach, sometimes adjusted for known differences in project complexity.
  2. Parametric Estimating: Uses an algorithm to calculate cost or duration based on historical data and project parameters (e.g. weight, distance, volume). It uses a statistical relationship between historical data and other variables (e.g. number of sites, kilometers of fiber) to calculate and estimate for activity parameters.
  3. Program Evolution and Review Technique (PERT), or Three-Point Estimating, is a step deeper than Parametric Estimating. Increases the accuracy of single-point estimates by considering estimation uncertainty and risk.

A detailed description and examples of these techniques can be found in the Primer on Critical Path Method & Estimation Techniques.

4.2.2 Task Sequencing

The last step to generate a preliminary Schedule is to sequence tasks, i.e. orchestrate tasks in a logical sequence that is compliant with NaaS Operators strategy, goals, and objectives and obtains the greatest efficiency given all project constraints.

This process involves the identification of relationships among the project activities having a clear view of predecessors and successors as well as lead or lag times between activities. Sequencing can be performed by using project management software or by using manual or automated techniques.

The use of Project Management software provides several advantages as it facilitates iterations and moreover, with the use of centralized/cloud tools, the Schedule & Plan can be made available to the entire organization making it easier to create awareness and focus efforts throughout the NaaS Operator areas.

Figure 3 shows a generic Schedule for a NaaS Initiative that can be customized by NaaS Operators to fit their own tasks, timelines, and logical sequence. To that end a set of guidelines and recommendations for customization are provided below as well as the associated High Level Schedule Template.

Figure 3. Generic NaaS High Level Schedule.

  • Build Project Plan. This task instantiates the plan described in this document and serves as a guide for all the other tasks. Three weeks can be considered to apply several iterations and gain acceptance and alignment among the core/executive team.
  • Commercial Agreements. Various commercial agreements should be in place before final definition of the Network Architecture, as these agreements may guide some of the decisions made for architecture. This can be a lengthy process that requires engagement with partners, analysis, negotiations and closing agreements. In consequence adequate duration must be considered (12 weeks are shown, but no less than 2 months must be allocated)
  • Spectrum Acquisition. This is one of the critical tasks (identified in yellow in Figure 3). Without spectrum, the NaaS Operator cannot offer any end user services. In some cases this might be de-risked through previous agreements with partner MNOs. However, if thats not the case, duration needs to be carefully adjusted considering the acquisition approach and the associated delays. Without spectrum secured, Network Architecture nor Network HLD can be finalized.
  • Headquarters / Remote Head Office Set-up. Formal activities of the NaaS Operator can commence remotely but at some point, a centralized facility is required to host resources and operations of the NaaS Operator, even if it is just a Remote Head Office for warehousing and as a central point of reference. This task and the associated milestone should be completed before completing hiring and staffing, although its desirable to have it before. Time will depend on the acquisition approach: the generic plan shows the case for a leased facilities without furniture for which 6 weeks is reasonable; If leasing a managed facility time to completion might be less than 6 weeks; and if building a facility it will take longer.
  • Hiring & Staffing. This task is usually active for several months. However, the complete project relies on internal staff; thus, the minimal team should be onboarded before the Network HLD starts and the full team should be onboarded in the worst case before Training & Preparation begins.
  • Other Resource Acquisition. This is a miscellaneous task to get any other resource not considered as part of the vendor selection and onboarding. Timelines will depend on specific resources required and the related tasks.
  • Network Architecture Definition. This is an important task as it provides inputs to the overall design and deployment activities. It depends on Spectrum acquisition and commercial agreements to be completed; and in turn, if affects Network HLD since this task cannot be completed if the Architecture is not finalized. Duration may vary from 2 weeks to several months depending when it starts, predecessor tasks and expertise of the NaaS Operator to perform the required analysis.
  • Network HLD. The duration of this task will vary based on the number of sites to be deployed and the amount of resources available. Number of sites per week will further depend on the complexity of the design and level of detail. As a general criterion, HLDs will take less time than LLDs. Without HLD, equipment vendor selection and site acquisition cannot start.
  • Vendor Selection & Onboarding. This task contemplates vendor selection and onboarding of various vendors such as construction company, deployment contractor and equipment vendors. Minimum duration is 4 to 6 weeks to allow for an adequate engagement, analysis, selection and onboarding of vendors. However, it might be extended. Without final vendor selection, network LLDs cannot be finalized and in consequence deployment cannot start.
  • Site Acquisition & Permits. Another critical task that carries a high risk as it cannot be fully controlled by the NaaS Operator and is delay-prone. Depending on the acquisition approach, risk can be reduced. However, for greenfield sites a minimum of 4 months should be considered being aware that this could take longer. For low-risk sites or those that are acquired before, Network LLD can start and be completed after the last site is acquired.
  • Network LLD. Similar to HLD, the duration of this task will vary based on the number of sites and the amount of resources available. However, the time to generate LLDs will be significantly higher than HLDs due to the level of detail required. Nevertheless, actual duration may get extended due to Site Acquisition and other commercial or administrative issues.
  • Hardware Delivery. Network equipment is usually shipped from abroad. Thus, delivery time, after purchase orders are issued must contemplate production (as required) and shipping. This is usually no less than 45 days. However, it can be de-risked based on lead times informed by the vendor during RFI/RFP and contracting process.
  • Training & Preparation. While waiting for hardware all the preparation tasks for deployment can be done. Ideally, this task is synchronized to end at the same time that hardware is delivered so deployment can start without delays.
  • Site & Fiber Construction. Whenever any sites or fiber infrastructure is build, this task should be considered, allocating adequate times for the number of sites and resources for construction. Deployment of a specific site cannot start until construction of civil and fiber infrastructure is finalized.
  • Site Deployment. Site deployment starts once the hardware has been delivered and any construction of site or fiber infrastructure is finalized. For multi-site deployments deployment can start with existing sites before the construction works are over in all the sites, as shown in the Figure. As a rule of thumb for timelines, it can be assumed that one week per site per crew to deploy sites.

4.2.3 Schedule Validation

The preliminary Schedule may need to go through several iterations before final acceptance from NaaS stakeholders. These iterations may be triggered by specific constraints, change requests from the executive team, new information related to timelines or status and other factors.

In addition, for each iteration, the Schedule must be validated with respect to the Strategy, achievement of objectives and the Resource Plan. To validate the Schedule the NaaS Operator must answer the following questions:

  • Is the Schedule accomplishing the NaaS strategic objectives?
  • Are timelines acceptable?
  • Are timelines feasible?
  • Are there any flaws in the logical sequence?
  • Is there any conflicts with resource plan such as overallocation or allocating tasks before resources are acquired?
  • Are there any execution risks, and if, so are these risks acceptable and manageable?

If an issue is detected, the following techniques and approaches can be put into practice to update the Schedule and/or the Resource Plan.

  • Resource Leveling. When shared or critically required resources are available only at certain times or in limited quantities or are over-allocated, NaaS Operators may adjust the start and finish dates of activities. This technique is often called Resource Leveling.
  • Smooth activities. Say there is an activity that takes 15 days but has a float of 15 more days before other successor activities may start. In this case, the resources of this activity may reduce their effort to 50% of the planned levels, smoothing and lengthening this activity up to 30 days.
  • Fast Tracking. Through a review of the Schedule, NaaS Operator looks for sequential activities that may be done in parallel to some degree.
  • Crashing. For activities which duration depend on the number of resources dedicated to them (e.g. building a wall with 3 people vs dedicating 6 people to it). Therefore, the team may find such activities in the critical path, and apply additional resources to it, to shorten the global project duration.

5 Communications Plan

The Communications Plan is the component of the High Level Project Plan that documents how communications across the NaaS Organization will be structured, implemented and monitored. As mentioned in Section 2.2.4, a well designed and implemented Communications Plan avoids delays and rework due to miscommunication and lack of coordination which is critical for NaaS Operators due to the high cost of re-work and financial constraints that are faced in rural environments.

Sections below present an examination of the required components for an effective Communications Plan, providing guidance for the NaaS Operator to customize reporting mechanisms, communication channels and document management rules and systems.

5.1 Meetings & Reporting Mechanisms

Meetings and reports are valuable means to communicate information and coordinate efforts in any organization. However, they can be a source of overhead reducing efficiency to the overall activities. For this reason, it is important to determine which information and when needs to be communicated through meetings and/or reports.

In the case of meetings, the NaaS Operator can establish periodic and ad-hoc meetings, in addition to strategic governance meetings defined in the Strategy & Scope Module. Periodic meetings have a concrete purpose that require recurrent synchronization within a specific set of stakeholders, and are scheduled in advance with a specific frequency.

Purpose of periodic meetings must be established by the NaaS Operator based on stakeholder needs, management and coordination requirements. Common purposes of periodic meetings are:

  • Project Planning / Steering. Formal meetings with the management team and key stakeholders to analyze and make planning decisions that affect an specific team or project.
  • Status/Progress Review. Formal meetings to report and evaluate current status and progress of ongoing tasks and activities.
  • Financial & Network Performance Meetings. Formal meetings to assess business and/or network performance based on KPIs and other standardized metrics and reports.
  • Stand-up Check points. Informal and quick meetings to brief project members of status of ongoing and upcoming activities that require coordination between team members.

Frequency of these meetings will vary based on purpose and the need for collaboration / assessment of the aspects mentioned above. It can be daily (mainly for stand-up check points), weekly, bi-weekly, monthly, or quarterly.

On the other side, ad-hoc meetings are asynchronous in nature and are created based on a specific communication need., such as team coordination, discussion of issues and changes, respond to contingencies (war rooms) and to report results from a major achievement or milestone.

In a similar way to meetings, reports can be scheduled with certain frequency or requested on-demand. Reporting takes time so its important to define efficient methods of reporting to reduce overhead to a minimum, and implement as much as possible ways to automate reports and use portals / cloud applications that allow anyone to dive in the information on their own. Based on the specific requirement reports can be email updates, Project Management assets, Dashboards & Tracking Sheets, slide presentations or spreadsheet reports.

The NaaS Operator must take into account that working in a rapidly changing environment requires communicating these changes more frequently and efficiently. Thus, frequent team checkpoints and updates are required. A set of meeting and reporting rules must be defined by the NaaS Operator. To that end, they can use the Communications Plan Template and the following recommendations:

  • Minimize overhead by defining the least number of meetings and reports that ensure adequate coordination, avoiding long meetings and defining simple reporting formats.
  • Include stand-up meetings as a key mechanism for team synchronization and coordination, especially for inter-departmental tasks.
  • Promote asynchronous reporting through a push model that allows the generator of the report to notify changes and updates without having to meet with the full set of stakeholders. This also minimizes interruption of actual work.
  • Allocate adequate time for meetings to avoid re-scheduling
  • Establish purpose and agenda for every meeting and send it ahead of time to create interest, allow for preparation and ensure engagement.
  • Involve only the required stakeholders for each meeting, avoiding crowded meetings with inactive or uninformed participants. In particular, it is important to minimize the number of people from the same team going to large meetings. Large meetings are only recommended if they are used to share key information across the company.
  • To close a meeting, sum-up agreements and next steps. In addition, one of the assistants should be designated as the note taker to share the overall notes and agreements with the participants.

5.2 Communication Channels

Communication channel plays a key role in how the information is received, processed and utilized by the recipients. Thus, the Communications Plan must establish communication channels to be used across the NaaS Organization, specifying in addition, under which circumstances one of them should be preferred.

Table 4 presents a comparison of typical communication channels which may be used by NaaS Operators to identify communication channels to be considered for their Initiative.

Criteria

Face-to-Face / Meetings

Email

Phone

Messaging apps

Video Conferencing

Enterprise Communication Platforms

Visual/Non-verbal communication

File Sharing

One-to-many conversations

One-on-one conversations

Mobility

Voice

Instantaneous notification

N/A

Cross-platform support

N/A

Table 4. Communication Channels

Once communication channels are established, the NaaS Operator may define simple rules to prioritize them based on the following criteria:

  • Formality. Phone, face-to-face meetings or email are good candidates for formal communications.
  • Urgency/Criticality. For urgent and critical matters phone or messaging apps are often preferred.
  • Collaboration. Depending on the degree and type of collaboration face-to-face meetings, video calls, enterprise communication platforms or messaging apps may be used.
  • Complexity. For complex or hard-to-explain issues face-to-face meetings are the golden choice. However, if thats not possible, video calls should be considered.

In any case it is important that the NaaS Operator promotes direct communication among personnel from all the areas and departments while collaborating, avoiding silos and setting apart hierarchies.

The Communications Plan Template can be utilized by NaaS Operators to document Communication Channels and utilization criteria.

5.3 Document Management

Document management is the set of processes and techniques to create, track and store digital or digitized documents. Usually, a Document Management System (DMS) is used to classify, retain, and protect documents, supporting versioning, collaboration and workflow automation.

For NaaS Operators is imperative to implement some sort of Document Management System, as it offers the following benefits:

  • Centralized Storage that prevents loss of consistency and integrity of information.
  • Document distribution and access management that ensures required information is opportunely accessed by authorized users only.
  • Fast and easy information retrieval that avoids delays on any tasks requiring documents or files stored in the system
  • Provide high-level information security.
  • Collaboration that enables multiple users to view and modify documents at the same time while maintaining consistency, allowing for an easy way to comment and update documents while maintaining consistency
  • Versioning that lets users retrieve previous versions of a document or continue working from a selected point.

Even with the cost associated with these tools, the value that is obtained from them, easily makes up for this cost. To have a simple example, inconsistency in versions can create misinformation, compatibility and lack of coordination that implies loss of time and money which can be solved through the use of these systems.

Table 5 provides a comparison of common Document Management Systems that can be considered by NaaS Operators for implementation in their Initiatives. In addition, small financially constrained NaaS Operators could use free versions of Drive, Dropbox or OneDrive to perform Document Management by sacrificing robustness, support and some integration, workflow and collaboration features.

Criteria

SharePoint

(Microsoft 365)

Alfresco

Onlyoffice

Drive1

(G-Suite)

Dropbox1

Storage Space

1 TB

Flexible

500 GB

Unlimited

Unlimited

Versioning

Workflows

Integration with other apps and systems

Office, ERP, BI, Project Management, Ticketing

ERP, email, cloud storage and Google Docs

Cloud storage

Office 365, Google Docs, Others through APIs

Office 365

Online editing

Device Synchronization

Deployment Model

Cloud

Cloud, On-premise, Hybrid

Cloud, On-premise, Hybrid

Cloud

Cloud

Cost

$12.5/user/ month

on-premise: free

Cloud / Hybrid: request quote

Cloud: as low as $1/user/ month
On-premise: $1200 per server (lifetime)

$12 /user/ month

$20 /user/ month

Note 1. Unlimited storage available starting at 3 users for Dropbox and 5 users for Drive

Table 5. Document Management System Comparison.

Selection and implementation of a DMS is just the first part of the work. For NaaS Operators is important not only to have a Document Management System, but to reinforce best practices around it. The following recommendations are provided as a starting point:

  • Enforce usage of the DMS avoiding local edition and distribution of shared documents
  • Promote collaboration through the DMS keeping stable versions of shared documents and information
  • Establish adequate permissions to access, edit and share documents
  • Optionally, develop workflows that allow to notify changes and request inputs for document preparation and completion as part of organizational processes.

Finally, the selection of the DMS and general rules for its usage can be registered by the NaaS Operator through the Communications Plan Template.


 
 

1. RAN – Introduction

The RAN Architecture module provides the NaaS Operator with principles and guidelines that, together, ensure the definition of a RAN Architecture that complies with its service and business requirements.

In addition, an end-to-end process is defined considering a range of implementation options for base stations to be adopted by the NaaS Operator for an optimal RAN Architecture definition.

Through the use of this module, NaaS Operators will be able to generate a Reference RAN Architecture as well as a detailed list of engineering guidelines and equipment requirements to guide subsequent design and deployment tasks to achieve business targets.

1.1 Module Objectives

This module will enable a NaaS Operator to define all architectural aspects regarding the Radio Access Network (RAN). The specific objectives of this module are to:

  1. Provide a general overview and fundamentals of the architectural options of the RAN in an LTE System.
  2. Provide guidance for the definition of the RAN Architecture that matches with NaaS Operators requirements, capabilities and business model.
  3. Enable operators to define engineering guidelines that will serve the RAN design team to generate compliant designs.
  4. Provide guidelines to generate requirements and specifications that can serve to evaluate and select vendors and specific equipment to be part of the RAN solution.

1.2. Module Framework

The NaaS PlayBook Framework shown in Figure 1 displays the PlayBook Modules and their relationship to the RAN Architecture.

The Strategic Plan & Scope provides business context for the NaaS Operator and serves as a guideline for the definition of the High-Level Network Architecture.

The RAN Architecture module is within the High-Level Network Architecture stream. The generated output of this module will serve as a starting point, scenarios, and constraints definition for the Network Design Modules.

Figure 2 presents a functional view of the RAN Architecture module where the main functional components are exhibited. Critical module inputs are further described and examined in section 2.2. In addition, Section 3 describes aspects for the RAN Architecture definition and organizes them into an end-to-end process to facilitate decision making regarding each component.

Figure 1. Network Deployment and Management Framework.

Figure 2. RAN Architecture Module Functional view.

2. Architecture Fundamentals

General overview of the baseline concepts of the RAN Architecture.

2.1 LTE RAN Architecture & Main challenges

LTE Network is composed by three major elements as shown in Figure 3:

Figure 3. High level view of LTE overall network.

  • Radio Access Network (RAN) Is a set of radio base stations that serve as access points to the LTE Network. Each base station (BTS) provides radio resources to the user equipment (UE) to access network services. In an LTE network, a BTS is also called eNodeB.
  • Transport Network Provides interconnection from the RAN to the Core Network. In some cases, it also connects different parts of the RAN. This network is composed of various elements depending on the selected transport solution: fiber, microwave, or satellite.
  • Mobile Core (Evolved Packet Core) Core Network which manages user sessions including authentication, mobility, and access to external networks (e.g. Internet). Some elements composing the Mobile Core Network are: Mobility Management Entity (MME), Home Subscriber Server (HSS), Serving Gateway (S-GW), Packet Data Network Gateway (P-GW) among others.

Further details of the overall LTE System Architecture will be presented in the Primer on LTE RAN Architecture Basics.

RAN can be composed by one or more base stations (BTS) distributed over the coverage region. Basic components of a Radio Base station are: Antenna, Radio Frequency (RF) unit (commonly known as Radio Remote Unit or RRU) and the Baseband unit (BBU). These components can be arranged in a couple of ways (configurations) depending on the location of the BBU versus the RRUs and antennas. These configurations are classified into Distributed RAN and Centralized RAN.

Distributed RAN (D-RAN) Architecture

This option reflects the most common architecture in which each BTS is composed by one BBU and one or more RRUs and their corresponding antenna. Interface between BBU and RRUs is known as fronthaul. Each BTS has direct communication to the Core Network via the Transport Network. Figure 4 displays the distribution for a D-RAN.

Figure 4. Basic concept of D-RAN.

Advances in radio design have led to the incorporation of the BBU and an RRU in a single box (all-in-one equipment) located closer to the antenna on the tower or pole. This means savings in both space and energy consumption, reducing CAPEX compared to the traditional BBU+RRU approach. This all-in-one (AiO) approach is a solution commonly used for small cell sites.

Centralized RAN (C-RAN) Architecture

Is an evolution to D-RAN in which every base station is divided into two parts: multiple RRUs and antennas remotely located on each site; each of them managed by a single centralized BBU. Figure 5 displays basic concepts of C-RAN.

Figure 5. Basic concept of C-RAN.

Communication between RRUs and centralized BBU is done via a so-called fronthaul network. Fronthaul network represents a set of elements with strict capacity, latency, and synchronization requirements. Usually, these requirements make necessary the deployment of fiber optic to connect the remote site with the centralized BBUs; in addition, distance between the centralized and the remote sites is constrained as well (Maximum distance between 10 and 40km). Further detail about the Fronthaul Network is provided in the Primer on Cloud RAN Basics.

C-RAN allows NaaS Operator to maintain or increase the number of network access points (RRUs), while keeping (almost) untouched the central BBU. This operating mode is often referred to as the central BBU and is seen as a pool of radio resources for the remote RRUs. C-RAN also allows functionalities that improve network performance by coordinating cells from different BTS. The Coordinated multipoint (CoMP) functionality is a transmission method in which the reception from different cells is tightly coordinated. This results in higher system capacity and improved data rates.

The centralization of a BBU simplifies radio resource management in areas with high concentrations of network users (such as populated towns, transportation stations, large events, etc.) that can put high stress on the base station that serves the area. In these scenarios, the action of adding more D-RAN type sites can increase cost and can lead to signal interference if the base stations are not carefully coordinated.

Although the need of a fronthaul transport network might appear to increase the total cost of ownership (TCO) of the RAN, it has been proved that that a centralized RAN reduces both the CAPEX and the OPEX compared with a traditional RAN.

In addition, some network suppliers offer the possibility of deploying virtualized versions of the physical BBU equipment as Virtual Network Functions (VNFs) to be deployed on a data center together with VNFs corresponding to different network components such as core network components. Virtualization is a technology that is gaining strong interest and popularity in 5G standards and networks. In fact, 5G is a Cloud-native solution that considers must of is components to operate in cloud environments, introducing flexibility to use the same equipment (like data centers) to host multiple virtual elements instead of current function specific equipment.

Deeper insight into the differences between D-RAN and C-RAN and how the integration of cloud RAN is performed will be given in the Primer on Cloud RAN Basics.

D-RAN vs C-RAN Cost Efficiency

C-RAN pools BBU resources, eliminating the cooling and space requirements at the cell site reducing both CAPEX and OPEX due to smaller footprint, lower electricity consumption and fewer maintenance visits.

However, as mentioned before, the use of a C-RAN architecture requires the introduction of a fiber network for the fronthaul transport network. In rural environments, this often represents an investment that is bigger than the savings introduced by the baseband pooling functionality and CAPEX/OPEX savings. In addition, in order to centralized BBUs, the distance between each cell site and the centralized BBU shall not be higher than 40 km.

Thus, in most rural scenarios D-RAN is the best option due to cost trade-offs and distance limitation. C-RAN could be selected in specific urban-like scenarios where the traffic demand is high and concentrated in small areas, ideally where fiber is already available.

Table 1 provides a comparison between C-RAN and D-RAN.

Table 1. Comparison of D-RAN and C-RAN Architectures.

2.2 Radio Access Network Architecture Design Inputs Description

Table 2 presents a description of module input data and candidate sources required for a RAN Architecture definition. Furthermore, the impact of each input on the design process is examined.

Table 2. Module Input Data.

2.3 LTE RAN Deployment Scenarios

A deployment scenario for the Radio Access Network refers to the available options the NaaS Operator can select to build the base stations. These options depend on NaaS Operator initial conditions and resources as well as service requirements.

While initial conditions determine how the base station is going to be built (as a Greenfield or as an Overlay site), service requirements impact on the type of the BTS (Macro Cell or Small Cell site).

Greenfield & Overlay Scenarios

RAN scenarios can be divided into two main categories depending on NaaS Operator initial conditions:

  • Greenfield The greenfield scenario means building the network without the support of existing sites or telecom infrastructure (such as existing towers). In this case, the design process (RAN HLD Module) has to work together with a site acquisition and construction process.
  • Overlay This scenario refers to an existing operator that wants to deploy a new Radio Access Technology (RAT) on top of its existing network. The most frequent case is that the operator has an existing network or sites and uses a new RAT (such as LTE) for capacity expansion.

Both scenarios have their own specifics and challenges to meet business and technical requirements. Deeper insight into this specifics and challenges will be covered in the RAN HLD and RAN LLD Modules.

Macro Cell & Small Cell Sites

Depending on coverage and capacity requirements as well as on geodemographic characteristics of the coverage region, there two main types of BTS that can implemented:

  • Macro Cell Site Macro Cell refers to a BTS capable of providing coverage for a large area (~5-10 km, depending on terrain conditions and selected frequency). This wide coverage is achieved due to high transmitter power (up to 40W per antenna port, depending on the equipment supplier) and high towers (up to 40 m and higher) and which are Macro Cell Site differentiators. They also require large infrastructure investment regarding installation (tower) and power supply. This makes Macro Cell Sites to be a high CAPEX option.
  • Small Cell Site Small Cells are low powered (usually up to 10W) BTS that does not require large infrastructure (usually installed on lamp poles or small towers (up to 10m)) for their operation. These characteristics make Small Cells achieve a smaller coverage range (~2-5km) compared to a macro cell range. In consequence, small cells are ideal to be placed in hotspots traffic areas to enhance networks capacity or to fill coverage holes left by Macro Cell Sites. However, in rural environments, small cells are usually deployed in regions where the deployment of a Macro Cell site is not cost-effective due to geodemographic and traffic characteristics of the region.

Figure 6 displays the coverage difference between a macro cell site and a small cell site.

Figure 6. Macro Cell Site vs Small Cell Site coverage difference.

Both BTS types provide radio access coverage but in quite different way making each more effective in different situations. Analysis for the selection of one type above the other will be detailed in the RAN HLD Module.

3. RAN Architecture Definition Process

Figure 7 displays the end-to-end process for the definition of the RAN Architecture.

A close up of a device Description automatically generated

Figure 7. RAN Architecture E2E process.

First step of the definition process is to define the RF Specifications which contemplate the selection of the operating band, guardband definition (in the overlay case) and the maximum allowed transmitter power for all BTSs.

RAN Topology & Scenario Definition section will guide the selection of the BTS architecture, the deployment scenarios, the BTS type(s) (Macro cell or Small cell) to be considered. It also provides guidance for the evaluation of a Demarcation point between the RAN and the core network in case NaaS Operator is not deploying its own core network.

Functional Requirements will define the required network functionalities based on RAN Sharing schemes and VoLTE if it applies.

Finally, the Engineering Guidelines Definition section will provide the NaaS Operator instructions on how to define engineering rules to be considered during the design process.

3.1 RF Specifications

BTS operation is conditioned to multiple parameters. These parameters are:

  • Operating frequency band
  • Guardband requirements (for overlay scenarios)
  • Duplexing Scheme (which is associated to the operating band)
  • Transmission power

These parameters must be selected based on NaaS Operator service requirements, available resources, and countrys policies and regulations.

In this section, guidance for the selection of the mentioned parameters is provided.

3.1.1 Spectrum Definition

Electromagnetic spectrum is a limited resource that is managed by a regulatory body in each country. For this reason, the selection of an appropriate spectrum band and carrier bandwidth depends on factors such as the country’s regulatory policies, spectrum fees, available spectrum chunks, MNO partners spectrum position and unlicensed spectrum conditions.

In addition, for Overlay scenarios it is necessary to set guard bands between LTE and existing 2G and 3G technologies in order to avoid interference between the co-located systems and a degradation in the service.

Band Selection

NaaS Operator can select its operating band based on one of the following considerations:

  1. Re-farming (OVERLAY SCENARIO): LTE system can operate over existing 2G GSM and 3G UMTS bands (e.g. Band 3: 1800MHz or Band 8: 900MHz) in overlay scenarios. As a result, operators can re-farm parts of the spectrum currently used for 2G or 3G voice and data services to support LTE. Spectrum re-farming is a cost-effective way to increase capacity in existing networks without the need to bid for new spectrum and by reutilizing infrastructure to overlay LTE. However, the only way re-farming is feasible is when the NaaS Operator or its partners/customers have enough contiguous spectrum (this means that the operating frequency band has enough spectrum left to insert a new carrier) to allow simultaneous operation of two or three technologies in a frequency band.
  2. Leverage from partners spectrum: NaaS Operator can use the spectrum previously assigned to its partner (in case partner has already been assigned a 4G band).
  3. Use an Unlicensed band: Electromagnetic spectrum is a scarce resource that is regulated and assigned by each country governmental institutions. However, the utilization of unlicensed spectrum in the 5GHz range has been gaining interest in past years because it allows an operator to deploy LTE services without the necessity of bidding for a licensed spectrum. On the other hand, the use of unlicensed spectrum imposes transmission power limitations, which limits coverage area of the BTS. In addition, NaaS Operator needs to verify if other operations are using unlicensed spectrum in the area, as this may cause a considerable amount of interference. NaaS Operator can find additional information about unlicensed bands for LTE in the Primer for LTE in Unlicensed Spectrum. NaaS Operator must be aware that the operation of LTE in unlicensed spectrum may not be supported by all user devices.
  4. Get Spectrum from government: NaaS Operator can bid to be assigned an LTE band following country regulations. Moreover, some countries around the world have been interested in facilitating unused spectrum to service providers that are willing to bring mobile services to uncovered populations (see Rhizomatica initiative). NaaS Operator can see if this same principle applies in its own environment.

Spectrum bands differ from each other in terms of their usage (for coverage or for capacity), available bandwidth, clutter type (Rural Urban) and if they are paired (block of spectrum in a lower frequency band and an associated block of spectrum in an upper frequency band) or not.

From a usage point of view, low frequencies (< 1GHz) provide extended network coverage at lower cost as fewer base stations are required to achieve greater geographic coverage. On the other hand, Higher frequencies (>1 GHz) are primarily used to cover urban and suburban areas where data traffic is dense and substantial network capacity is required (higher spectrum bands normally provide larger bandwidths which allow for larger transmission rates). Figure 8 shows spectrum bands usage from a coverage and capacity point of view.

Figure 8. Coverage and Capacity characteristics of spectrum bands.

NaaS Operator can leverage on Table 3 to choose a band that best fits its business strategy.

Table 3. LTE bands main characteristics.

LTE spectrum can operate in a wide range of bandwidths: 1.4, 3, 5, 10, 15, or 20 MHz.

If the NaaS Operator has the opportunity to select an specific LTE spectrum band for a rural network deployment, it is recommended to aim for a low band (below 1GHz) and a bandwidth that covers its present and future necessities in terms of coverage and capacity.

Cell bandwidth of 20MHz is often used in urban and suburban environments in which capacity is the focus. However, having a 10MHz bandwidth can be a good tradeoff between coverage and capacity performance. A 5MHz bandwidth can be useful to bring connectivity to an uncovered population but will be limited in terms of capacity performance. Having large bandwidths (e.g. higher than 20MHz) introduces the possibility for a capacity expansion using a second carrier.

For instance, in recent country-wide rural projects, most mobile network operators (MNOs) and NaaS Operators have selected Band 28 700MHz and 10MHz cell bandwidth as this combination has proven to perform good to provide rural coverage.

Guard band Requirements (OVERLAY SCENARIO)

One of the important aspects when introducing the LTE system on top of existing systems like 2G-GSM or 3G-UMTS is the guard band requirement. More specifically, when LTE is co-located with other technologies in the same site where nearby antennas or sharing scenarios are adopted then a guard band needs to be maintained to avoid interference between collocated systems.

The guard band requirement can be summarized as follows:

    • In the case of co-located LTE and GSM operating on the same band, then a single carrier of GSM 200 kHz is enough to avoid interference between the two systems as shown in Figure 9.

Figure 9. Guard band requirement for collocated LTE and GSM systems in the same band.

  • In the case of LTE co-located with another system in a different band (e.g. LTE at 2.6 GHz and GSM 1800 MHz) then no guard band is needed.
  • In the case of co-located LTE and UMTS operating on the same band, no guard band is required, however, both systems need to deploy a strict filter on the RF power amplifiers to avoid interference.
  • In the case of FDD-LTE co-located with TDD-LTE, then half of the channel BW (whatever is higher between FDD and TDD) is needed as guard band.
  • In the case where LTE is co-located with another LTE system (from another operator) in the same band, there is minimum to no interference due to the orthogonal nature of LTE access scheme. For this reason, there is no need for guard bands if adjacent technologies are FDD-LTE.

Table 4 summarizes the guard band requirement for LTE with different collocated technologies.

Co-existing systems

Required Guardband

LTE band X + GSM band X

0.2 MHz (One GSM carrier)

LTE band X + LTE band Y

0 MHz

LTE band X + UMTS band X

0 MHz (RRU must implement a strict RF filter to avoid interference)

LTE-FDD + LTE-TDD

Half of the channel BW (the higher one)

LTE band X + LTE band X

0 MHz

Table 4. Summary of guard band requirements.

3.1.2 LTE Duplexing Scheme Selection

LTE has two schemes for resource sharing and achieve two-way communication (between the eNodeB and the UE): Frequency Division Duplexing (FDD) and Time Division Duplexing (TDD). FDD scheme uses one frequency for downlink (DL) and another frequency for uplink (UL); allowing simultaneous communication between BTS and UE. On the other hand, in TDD scheme both BTS and UE use the same frequency for their transmissions, preventing simultaneous communication and implementing a dedicated time slot for transmission (TX) and for reception (RX). Figure 10 shows the operation of FDD and TDD schemes.

Figure 10. LTE Duplexing Schemes.

NaaS Operator must be aware that the Duplexing scheme also depends on the operating band. In case the selected (available or allocated) frequency band differs from the chosen duplexing scheme, NaaS Operator will be tied to work with the scheme associated to that band.

Table 5 shows a comparison of the FDD and TDD schemes from transmission rate, coverage, and CAPEX performance points of view. NaaS Operator can use this criterion to choose the scheme that best fits its service requirements.

Performance Item

FDD – LTE

TDD – LTE

Transmission Rate Performance

FDD allows higher throughputs.

Transmission rates cannot be achieved at similar distances when compared to FDD. However, it allows the implementation of advanced antenna techniques such as beamforming.

Site Count

Since FDD achieves cell edge rates at farther distances, the number of sites required to serve a given area is lower than TDD.

In a system comparison using the same frequency band, the TDD system requires 31% more base stations than FDD when using a 1:1 (same number of RX and TX slots) and 65% more base stations when using a 2:1 (twice the amount of RX than TX slots) due to capacity constraints.

Spectrum Arrangement

FDD operation requires twice more spectrum for a single LTE carrier for a given channel bandwidth.

TDD can operate in an unpaired spectrum. In addition, it can also be suitable for asymmetric transmission demands (like fixed wireless access deployments).

CAPEX

The fact that FDD requires fewer BTSs to serve a specific area using the same frequency allows the NaaS Operator to incur in lower equipment and infrastructure associated costs.

On the other hand, FDD spectrum fees can make it a more expensive option.

Table 5. FDD vs TDD performance comparison.

3.1.3 Transmitter Power Definition

Together with the operating band, the maximum transmitter power (per antenna port) is one of the most important parameters to be considered previous to the design step. BTS transmission power must be set based on the following considerations:

    • BTS Type There are no formal rules to define the transmitter power of a BTS based on its Type (Macro Cell or Small Cell). Table 6 displays classification between BTS types based on the transmitter maximum power.

BTS Type

Radiated Power

Medium Range (Small Cell) BTS

Up to 10W (40 dBm)

Wide Area (Macro Cell) BTS

There is no upper limit for the radiated output power of a Wide Area (Macro Cell) BTS

Table 6. BTS classification based on transmitter power.

For the case of small cells, most equipment in the market can achieve 10W (40 dBm) as maximum transmission power. In MORAN scenarios (as will be discussed in section 3.3.1) a 10W small cell can allow the implementation of two independent 5W carriers (one per Mobile Operator). This rule should be always considered for small cell definition because increasing the power with small transmitter heights has diminishing returns.

For transmit power ranges above 10W, network design should consider a macro cell site for the design process. Maximum power can be set according to the following point.

    • Tower height: As will be detailed in the RAN HLD module, the higher the transmitter is from the ground, the wider the coverage area achieved by the BTS (tower height consideration does not have an important impact in small cells as these are usually located in smaller infrastructure). In this sense, NaaS Operator can use this principle to set higher transmitter power to those sites that will be placed on high towers in order to reach a wider coverage area. Table 7 shows an example of how to define maximum allowed transmitter power for macro cell sites according to transmitters height.

Transmitter Height

Maximum Allowed Transmit Power

< 10m

<= 10 W

From 10m to 20m

10W to 20 W

From 20m to 30m

20W to 30W

> 30m

>= 30W

Table 7. Tower height and Maximum Allowed Transmitter Power relation example.

  • Local and regional regulation NaaS Operator must consider that some regulations may exist in certain regions preventing high transmit powers to be used. These regulations are mandatory requirements to be followed by all operators in the region. For instance: maximum allowed transmitter power for a certain region can be set to 30W. In this case, NaaS Operator must arrange its Transmitters Height vs maximum allowed transmitter power considering the 30W limitation. NaaS Operator must be aware of any local and regional regulations.

3.2 RAN Topology & Scenario Definition

This section guides NaaS Operator on its way to standardize the topology that will be implemented on the RAN considering possible scenarios which include:

  • D-RAN vs C-RAN Architecture.
  • Greenfield and/or Overlay deployment.
  • Macro Cells and/or Small Cells
  • Use of a Demarcation Point between access and core network

3.2.1 D-RAN vs C-RAN Definition

As shown in Table 1, there are several tradeoffs that must be considered when selecting between D-RAN and C-RAN. In addition, this selection will condition several aspects of the overall network, mainly transport and core design and infrastructure.

Despite being a promising solution in terms of performance management and future-proof, C-RAN is feasible only in a reduced number of scenarios. Considering this, the decision process of selecting between D-RAN and C-RAN can be reduced to the following points.

  • Number of coverage areas: Coverage areas can be defined as towns or cities that will be covered. A C-RAN architecture wont be efficient for a low number of coverage areas (less than 3).
  • Separation between coverage areas: if there exists a separation greater that 40 km between coverage areas, C-RAN architecture is not feasible due to latency issues, making D-RAN the adequate choice.
  • Optical Fronthaul Network: C-RAN requires an optical fronthaul network which comes with a high investment. In addition, the length of the links must be below 40km for C-RAN sites to operate properly. C-RAN can only be considered if NaaS Operator can afford the construction of a fronthaul (fiber) network.

Methodology to select between a D-RAN or C-RAN architecture is shown in Figure 11:

A close up of a map Description automatically generated

Figure 11. Methodology to select between a D-RAN or C-RAN architecture.

3.2.2 Greenfield / Overlay Scenarios Identification

NaaS Operator can Identify its initial scenarios considering the following points:

  • Greenfield site definition No existing sites available.
  • Overlay site definition NaaS Operator already has an existing operating network in the desired region and all sites of the network have enough resources to support the addition of a new RAT on them.
  • Greenfield + Overlay NaaS Operator already has an existing operating network in the desired region, however, only some of these sites have enough resources (available space on tower and enough power supply for new equipment) to support the addition of a new RAT. The remaining sites will need to be considered Greenfield ones.

Figure 12 displays a methodology to identify which are the initial scenarios in a NaaS Operator environment. NaaS Operators network can be composed by a combination of Greenfield and Overlay sites if conditions are met.

A close up of a map Description automatically generated

Figure 12. Methodology to select between Greenfield and Overlay scenarios.

Existing sites must be able to accommodate new infrastructure (BBU, RRUs or antennas) to be considered as Overlay ones. This means that they must contain enough space on the towers and power sources to install and feed the new equipment. A site survey should be held in order to evaluate if existing sites are candidates for an overlay deployment. The RAN HLD Module provides a questionnaire for the evaluation of existing sites in order to define whether they can be used for an overlay or not. This evaluation must be held by a site survey to the desired site. Details for the site survey will be given in the Site Survey Module.

3.2.3 Macro / Small Cell Scenarios Definition

The NaaS Operator network can be composed of a combination of macro cells and small cells. Table 8 displays a technical comparison between Macro Cells and Small Cells.

 

Macro Cell BTS

Small Cell BTS

Main purpose

Coverage and capacity

Additional Capacity and cover small targeted areas.

Can be selected if Macro Cell BTS is too expensive.

Transmit Power

Up to 80W (depending on equipment suppliers capabilities)

Up to 10W

Cell Coverage

~5 10 km

~2 5 km

Site Solution

BBU + RRU + Antenna

All-in-One equipment.

Integrated BBU and RRUs (in some cases, antennas can also be integrated into the same box.

Antenna Configurations

2×2, 4×4 MIMO

2×2 MIMO

Installation options

Tall towers (10m 40m)

Rooftops

Small towers (up to 10m)

Lamp poles

Rooftops

Table 8. Macro Cell and Small Cell BTS technical comparison.

This decision-making process of selecting BTS Types must take into consideration several conditions ranging from technical, infrastructural and financial constraints to population distribution and traffic behavior. The combination of these conditions may raise the appearance of multiple scenarios that will vary from one region to another. Table 9 displays a scenario definition based on population classification.

Town Size

Population Density

Separation between Towns

Small

Low

High

Small

High

High / Low

Big

Low

High / Low

Big

High

Low

Table 9. Scenario definition.

Multiple scenarios may bring the necessity to consider a wide range of BTS Types and configurations.

However, in order to have a homogeneous network (which will simplify future processes like site catalog, network construction and maintenance, among others), NaaS Operator can reduce all possible BTS configurations into a few options. While having a low number of configurations might limit performance, it allows for a cost-effective and easy operation of the network.

Naas Operator can provide BTS configurations as shown in Table 10 to be considered for all its network. Design team will decide (during the HLD process) which option best suits a specific area.

BTS Type

Number of sectors per BTS

Tower Hight

Small Cells

– 1 sector (omni)

Indistinct

Macro Cells

– 1 sector (directional)

(Only for highway coverage)

Indistinct

– 2 sectors (directional)

10m to 20m

– 3 sectors (directional)

> 20m

Table 10. Number of sectors definition per BTS type.

3.2.4 Demarcation Point

Demarcation point in a network divides it into two or more sections (depending on the number of demarcation points). For instance, the S-GW acts as a demarcation point between the RAN and the core network.

Demarcation point in the RAN is required if the core network is not maintained/owned by the NaaS Operator. The Demarcation point allows to have control over RANs performance and to perform charging activities to its customers that are Mobile Operators. If the deployment of a mobile core is considered as part of the architecture, then, the demarcation point is the SGW or PDN Gateway. On the other hand, if there are no plans for a mobile core, then a demarcation point must be considered in the RAN.

The following are some functionalities introduced by a demarcation point in the RAN:

  • NaaS Operator Independence: Creates a logical separation between MNO and NaaS Operator network. This enables the deployment, configuration, operation, and monitoring of sites without depending on the MNO.
  • Security towards core network: Provides a security layer that protects the core network.

Some demarcation points that can be implemented in the RAN are:

    • RAN Gateway for traffic aggregation

This gateway (GW) improves management and performance operations in multi-RAT or multi-MNOs scenarios. Additional functionalities introduced by the RAN GW are:

      • RAN Homologation: Abstracts the complexity of multi-RAT solutions towards transport networks.
      • Multi-vendor Support: Connectivity with equipment from different network suppliers.
      • Connection with 3er parties: Facilitates interconnection of 3rd parties to the MNOs.

The RAN GW general concept is shown in Figure 13.

Figure 13. RAN Gateway concept.

    • IPsec Gateway

An IPsec GW can be introduced to provide IPsec tunneling termination towards RAN and towards core network, thus, improving network security for all parties in the network. IPsec GW concept is displayed in Figure 14.

Figure 14. IPsec Gateway concept.

NaaS Operator must be aware that demarcation point functionalities may vary from one equipment supplier to another.

3.3 Functional Requirements

This section will guide the NaaS Operator in the definition of the RAN functionalities required for the operation of the defined topology and service requirements defined as part of the NaaS Operator business strategy.

3.3.1 RAN Sharing Scenarios Definition

NaaS Operator central strategy is to enable business to MNOs in new areas without them having to invest on network infrastructure. Instead, MNOs can provide their network services using NaaS Operators infrastructure and network resources.

NaaS Operator can provide its RAN infrastructure to multiple MNOs based on the following sharing schemes:

  • Multi-Operator RAN (MORAN): This scheme allows to share BBUs between MNOs but not radio carriers (RRUs are not shared). This implies that every MNO will serve its subscribers in its own spectrum chunk.
  • Multi-Operator Core Network (MOCN): This scheme allows all elements in the RAN (BBUs and RRUs) to be shared. The difference is that MOCN allows radio carriers to be shared too which makes MOCN to be the most resource efficient. Moreover, MOCN can be deployed quickly as it only requires the feature to be configured in the RAN. This makes MOCN ideal as a solution for quick coverage expansion. By activating MOCN on an existing network, subscribers from a new MNO customer can get access to the NaaS Operator network. Mobile cores of the various MNOs using the NaaS Operator network are independent.
  • Roaming Agreements between NaaS Operator and MNO: In this scheme no RAN resources are shared. Instead, NaaS Operator owns the RAN and the core network and redirects traffic to MNOs core networks. In this case, the user might have to enable roaming on their phone to access the NaaS  Network.

NaaS Operator can use Table 11 to decide which sharing scheme best fits its strategic plans.

Table 11. RAN Sharing application schemes.

Figure 15 illustrates the implementation of each RAN sharing scheme.

Figure 15. Possible RAN Sharing Schemes.

Table 12 summarizes technical characteristics of each scheme.

Table 12. RAN Sharing schemes characteristics.

3.3.2 VoLTE Implementation Evaluation

Voice over LTE (VoLTE) is the 4G-LTE solution to provide voice services in an LTE network. VoLTE provides high-quality voice and video services. However, this section only applies if at strategy level the NaaS Operator has decided to provide voice services.

The implementation of VoLTE on a network brings the following requirements:

  • Integration of an IP Multimedia Subsystem (IMS) on the core structure as standard EPC is not able to support VoLTE on its own. Possible strategy is to reach an agreement with MNO(s) in order to redirect traffic from NaaS Operators core network to MNOs IMS or from NaaS Operators RAN to MNOs core network (which includes an IMS).
  • Additional RAN functionalities (that will be detailed in section 3.3.3) that are basic for the operation of VoLTE in the air interface.

These requirements represent an addition to the networks CAPEX, increasing NaaS Operator investment. Additionally, the configuration of VoLTE on a BTS means extra integration efforts.

However, NaaS Operator can avoid the implementation of VoLTE if LTE BTSs are deployed as an overlay in locations where existing 2G and/or 3G sites are already operating. Voice in overlay scenarios can be carried by the existing RATs using the Circuit Switched FallBack (CSFB) functionality which allows an user equipment (UE) connected to the LTE system to move to 2G-GSM or 3G-UMTS in order to complete the voice communication. Once the voice call is over, the UE returns to LTE. CSFB principle is shown in Figure 17.

Figure 16. Circuit Switched FallBack principle.

For these reasons, the implementation of the CSFB functionality is preferable from financial and sites configuration points of view. However, CSFB principle does not apply in cases where there are no existing sites operating with 2G and/or 3G. In other words, VoLTE is the only option to provide voice services on Greenfield Scenarios.

Figure 17 displays a strategic criterion to guide NaaS Operator in the decision-making process of deploying VoLTE:

A close up of a map Description automatically generated

Figure 17. VoLTE decision-making process.

3.3.3 Equipment Requirements Definition

Network equipment chosen for the deployment of the RAN must be able to support the selected architecture that has been established up to this point and will depend on the BTS Type: Macro Cell or Small Cell.

In addition, RAN features must be considered too in order to allow the implementation and operation of the defined RAN Architecture and functional requirements that have been established up to this point.

Table 13 summarizes equipment requirements and RAN features required based on defined RAN Architecture requirements.

NaaS Operator can use the RAN Equipment Requirements Template to create its own version.

Table 13. Equipment requirements based on RAN Architecture specifications.

3.4 Engineering Guidelines Definition

Engineering Guidelines are a set of rules that need to be followed during the High-Level Design (HLD) Process. Engineering Guidelines must be set according to the defined RAN Architecture and should cover the minimum requirements for the network to operate as desired. Network designer will establish site configurations based on the given rules. These rules cover Radio (coverage and capacity) as well as Infrastructure aspects of the network.

NaaS Operator can use the e the Engineering Guidelines Template to set the engineering rules to be considered during the HLD process. Recommended values will be given for each parameter as a starting point for the NaaS Operator.

3.4.1 Coverage & Capacity Guidelines Definition

Guidelines for the radio aspects of the network include parameters that must be considered during network dimensioning. Table 14 summarizes the required Coverage and Capacity parameters that must be set before the HLD process (additional parameters will be introduced and explained in the RAN HLD Module).

Table 14. Description of Coverage and Capacity engineering rules categories.

3.4.2 Infrastructure Guidelines Definition

Guidelines for the infrastructure aspects required to build site configurations during the HLD process. Table 15 summarizes the categories of the infrastructure engineering rules:

Table 15. Description of Infrastructure requirements categories.

4. RAN Architecture Report

A RAN Architecture report must be elaborated in order to provide a clear image of the overall network and engineering guidelines to the design team. Additionally, equipment requirements can be considered for the RFx process.

The report must cover all points and considerations that define each BTS in the RAN and its main functionalities. Such points are:

  • Reference RAN Architecture Defined overall RAN architecture.
    • RF Specification
      • Band and Guard band
      • Transmitter Power
    • RAN Topology and Scenarios
      • BTS Architecture and Type
      • Deployment Scenario
      • Demarcation Point
    • Functional Requirements
      • RAN Sharing Scheme
        • Voice Services Implementation
        • Equipment and Functionalities requirements
    • Engineering Guidelines
      • Coverage, capacity, and infrastructure rules to be considered during the design process.

NaaS Operator can use the RAN Architecture Report template as a reference to create its own version.

Annex I

Table 16 illustrates the supported bands Frequency Division Duplex (FDD) LTE while

Table 17 illustrates bands for Time Division Duplex (TDD) LTE according to the 3GPP.

Table 16. FDD-LTE 3GPP Bands.

Table 17. TDD-LTE 3GPP Bands.

 


 

1. Transport and IP Architecture Introduction

This Transport and IP Architecture module aims to provide a frame of reference and enable a NaaS operator to define the overall network solution that interconnects the radio access network (RAN) toward the mobile core and data centers, including the technical aspects of the IP networking solution.

This module provides an overview of the transport and IP network fundamentals focusing on required functional and operational aspects to support the LTE service. Moreover, it provides guidelines to translate these requirements into a transport and IP architecture that better suits NaaS operator capabilities and constraints.

The output of this module is the transport and IP reference architecture that includes the overall solution description, as well as a formal requirements specification to guide subsequent design and vendor selection phases.

1.1 Module Objectives

This module will enable the NaaS operator to define the reference architecture of the mobile transport and IP network. The specific objectives of this module are to:

  1. Provide an information base and fundamentals of the architectural elements of the transport and IP network.
  2. Guide the NaaS operator to determine the transport and IP architecture functional and operational requirements to provide an LTE service.
  3. Provide instruction to define and specify the most appropriate transport and IP network architecture according to the NaaS operators requirements, capabilities, and business model.
  4. Provide guidelines to generate a consolidated reference architecture and requirements specification for evaluation and vendor selection for the transport and IP network.

1.2 Module Framework

The NaaS playbook framework in Figure 1 describes the structure, interactions, and dependencies among disparate NaaS operator areas.

The Strategic Plan & Scope provides business context for the NaaS operator and serves as a guideline for the definition of the high-level network architecture, which in turn drives the strategic decisions to forthcoming phases.

The Transport & IP Network Architecture module is included within high level architecture and has direct relation with RAN and mobile core architecture as well as network monitoring architecture. The generated output of this module will serve as a required input for the network design modules.

Leon - Rey de la Selva

Figure 1 Module framework

The transport and IP network architecture functional view is shown in Figure 2, where the main components are depicted in addition to inputs that drive the definitions made throughout the development of the transport and IP reference architecture.

Figure 2 Transport and IP architecture framework

The model structure is divided into four sections. Section 1 provides an overview and fundamentals involved in the design of the transport and IP architecture. Section 2 focuses on the functional and operational requirements of the transport network. Section 3 discusses possible variations to implement the transport and IP network and methodologies to perform the selection that best suits NaaS operator conditions. Section 4 illustrates how to integrate previous elements into a formal reference architecture and requirements specification.

 

2 Transport and IP Network Architecture

General overview of transport and IP network architecture and main transport technologies to be implemented in the network. In addition, the particularities of the IP network are examined.

2.1 Transport Network Overview

In a mobile environment, the transport network interconnects disparate networks, including the RAN, data centers, and external networks. Figure 3 displays the architecture of a typical transport network.

Figure 3 Typical mobile network architecture

Mobile networks are ubiquitous and support a mix of traffic types originating from and terminating to mobile devices. All this traffic must be conveyed between the mobile cellular base stations through the transport network up to the mobile core.

For this reason, different segments exist on the transport network that for most of NaaS operators can be classified as: last-mile and aggregation level. Additionally, there might exist a connection provided by a third-party network, which can provide the last mile or aggregation. The network segments are mostly defined by the technology and topology used within it. The key network segment characteristics are:

  • The last-mile and aggregation segments provide the connectivity to the eNodeB at the cell sites and are predominantly based on three chain topologies built with microwave radios and fiber optic links.
  • Optionally, a third-party network can be used to transport traffic to the mobile core. A typical method to implement this scenario is presented when the NaaS operator doesnt have network infrastructure in the geographical zone. In nearly all cases, the third-party network is an IP/MPLS network.

Figure 4 displays the structure of the transport network.

Figure 4 Basic structure of a mobile transport network

The structure displayed in Figure 4 considers different physical technologies and topologies. Moving left to right, the last-mile connects a demarcation device, usually deployed at the cell site, to a first stage of traffic grooming and concentration. The aggregation network, in turn, further aggregates traffic, adapts any technology change and provides the hand-over point to a higher level of aggregation network.

The term physical connectivity is used to represent any technology that can be used to connect nodes. Common physical connectivity includes fiber optic, microwave, and satellite links. In addition, on top of the physical layer a networking layer is implemented that embraces all the possible logical architectures needed to steer LTE traffic and applications.

2.1.1 LTE Flattened Transport Architecture

Traditional cellular networks (e.g., 2G and 3G) are hierarchical in nature; that is, there is no direct communication between the base station and the core elements. Base stations communicate with their respective radio controllers that are in turn connected to the core elements.

Long-term evolution (LTE) defines a flat network architecture that eliminates the radio controllers. Thus, controller functions are redistributed to the core elements and the base stations (eNodeB). For this, two logical S1 interfaces between the eNodeB and the evolved packet core (EPC) are implemented: the S1-MME to the mobility management entity (MME) and the S1-U to the serving gateway (S-GW). MMEs handle authentication and signaling aspects of the mobility management control plane. S-GWs terminate the user plane traffic. Figure 5 illustrates the LTE network architecture:

Figure 5 LTE flattened transport architecture

Figure 5 illustrates an overview of the logical planes needed in the transport network. Various kinds of user and control planes are transported across the transport network. In addition, management traffic that consists of the data needed for support of the network elements FCAPS (fault, configuration, accounting, performance, security) must be considered. When synchronization is arranged by the transport network (e.g., precision time protocol (PTP)), these packets need to be delivered to the eNB from the timing server.

2.2 IP Network Overview

On top of the physical media, the connectivity required by LTE can be realized by using different service types. All these traffic types are forwarded from the last-mile through the aggregation tiers to the core.

The most common implementation scenario including last-mile/aggregation and third-party network segments is displayed in Figure 6:

Figure 6 Transport network scenario based on carrier Ethernet with L3VPNs

The scenarios presented in Figure 6 have been defined only on top of Ethernet (IEEE 802.3), as its the dominant layer 2 technology. The use of Ethernet interfaces has also been assumed for all base station and mobile core nodes. More detail on the required transport protocols can be found in Section 4.4.

In the third-party network segment, an IP/MPLS-based transport is used to implement a L3 VPN on top of Ethernet. Using a L3 VPN permits the NaaS operators nodes to communicate privately over the third-party network, even though it routes through multiple third-party elements. Detail on this implementation is out of scope for this module, as that is a service provided by the third-party network.

2.3 Transport Technology Overview

The implementation of 4G LTE imposes some requirements on transport networks, such as more network capacity and latency reduction. These requirements are better served through terrestrial technologies (fiber optic and microwave). But in rural areas this becomes a challenge because satellite transport is usually the only feasible technology. For this reason, the transport network infrastructure is an essential component of the NaaS operator network.

An overview of commonly used technologies in the transport network is presented in the following subsections.

2.3.1 Fiber Optic Technology

Fiber optic is the mainstay wired transport in mobile operator networks whenever its available because of its significant, inherent bandwidth carrying capability. Additionally, several additional techniques can be used to offset any bandwidth constraints and essentially render the fiber assets future-proof.

While fiber optic has tremendous operational capacity, its main limitation is the cost and logistics of deployment. Moreover, it can take several months to provision each cell site with fiber optic transport.

The fundamental aspects of fiber optic technology design are discussed in the Primer on Fiber Optic Technologies Principles.

2.3.2 Microwave Technology

Most operators rely heavily on microwave transport solutions in the 5 GHz 80 GHz bands (frequencies above 15GHz are not generally feasible in rural environments). Microwave is a low-cost option for mobile transport, as it can be deployed in a matter of days and support a range of up to several tens of kilometers.

The main limitation of microwave links is the requirement of line of sight (LOS) between transmitter and receiver. This represents an important constraint for its implementation, especially in rural areas due to steep conditions in the terrain. In addition, in many cases microwave requires a license to operate. In high frequencies, microwave links are subject to atmospheric effects or rain fade, which can attenuate the signal and limit its range.

The fundamental aspects of microwave technology design are discussed in the Primer on Microwave Technologies Principles.

2.3.3 Satellite Technology

Satellite technology is deployed in fringe network areas, usually in rural scenarios in emerging markets. Furthermore, satellite technology can be deployed as a temporary measure as an operator waits for regulatory microwave licenses to be approved. Satellite coverage is determined by satellite footprint and location-specific geographical and meteorological conditions.

The cost of satellite terminal equipment located at the base station is in line with microwave equipment. However, the OPEX burden generated by the satellite service fee can impact on the business case.

The fundamental aspects of satellite technology design are discussed in the Primer on Satellite Technologies Principles.

3 Functional & Operational Requirements

Examination of functional and operational characteristics of the transport and IP network, providing guidance and general criteria to establish requirements for network designers and solution providers.

3.1 Transport Network Planes

As stated in section 2.1.1, each eNB requires connectivity to the core elements and additionally for network management and potential synchronization. The subsequent sections describe the different planes that must be supported by the transport network. More detail on required transport protocols can be found in Section 4.4.

3.1.1 User plane

The user plane (U-plane) in the LTE transport network consists of the S1-U plane used to transport user data between the eNodeB and the S and P-GW using a general radio service tunneling protocol user (GTP-U). It also includes the X2-U plane used to provide connection to neighbor eNodeBs.

The protocol stack required to support the user plane is shown in Figure 7.

Figure 7 LTE user plane protocol stack

3.1.2 Control Plane

The control plane (C-plane) in the LTE transport network is comprised of the S1-MME traffic used to transfer signaling information between the eNodeB and MME. This information is used for S1 bearer management, mobility, and security handling, as well as for application signaling messages between the user equipment (UE) and MME. It also includes the X2-C plane used to provide connection to neighboring eNodeBs.

The protocol stack required to support the control plane is shown in Figure 8.

Figure 8 LTE control plane protocol stack

3.1.3 Management Plane

The management plane (M-plane) in the LTE transport network is the interface between the eNodeB and the O&M system. The M-plane consists of all data required to manage and monitor the status of the network elements. The volume of this traffic has to be considered during the capacity analysis and the frequency to collect these data.

The protocol stack required to support the management plane is shown in Figure 9.

Figure 9 LTE management plane protocol stack

3.1.4 Synchronization Plane

The synchronization plane (S-plane) in the LTE transport network is important to ensure proper functionality of the eNodeBs. The two types of synchronization methods for a mobile network are phase/time and frequency. More detail on the synchronization requirements is in Section 4.7.

However, the implementation of this plane can be optional if its decoupled from the transport network by using GNSS (global navigation satellite system) solutions at the eNB.

3.2 Availability & Resiliency

Most of the design considerations for the transport network architecture are constrained by network ability to meet the proper levels of availability (proportion of time a system is in a functional condition) that is supported by means of resiliency (ability to readily recover). Table 1 shows the downtime per year according to the availability value.

Availability

Downtime Per Year

99%

3.65 days

99.9%

8.76 hr

99.99%

52 min

99.999%

5 min

Table 1 Availability and corresponding downtime per year

NaaS operators must define different availability levels for the transport network according to network segment (last-mile/aggregation). Figure 10 displays typical values of availability and resiliency requirements across the transport segments.

Figure 10 Availability on each segment of transport network

As seen in Figure 10, availability requirements of the transport network are derived from the availability requirements of the end-user service. A typical requirement for the transport network in a rural environment could be an availability value of 99.95%. In most cases the availability is dominated by the last-mile link.

Recommended values for the various segments within the rural NaaS transport network are presented in Table 2.

Service Level

Availability (%)

Last-mile Link

99,9%

Aggregation Link

99,99%

3rd Party Network

99,99%

Satellite Link

99,9%

Table 2 Availability requirements definition example

3.3 Quality of Service Support

Quality of service (QoS) management is essential to distribute limited resources between different traffic classes. The concept of QoS used in LTE networks is based on classes, where each type of traffic (e.g., data, voice, signaling) is assigned a QoS class identifier (QCI) by the network to ensure adequate QoS for carrier traffic in the LTE network.

To allow separation of traffic in the transport network, translation of the radio QCI level into the transport QoS level is performed initially on the eNodeB. However, additional translations may be performed by intermediate transport devices along the path. For this reason, the QoS parameters in all the devices across the path must be correctly configured.

To allow traffic separation, the IETF introduced the notion of differentiated services (DiffServ) classes as a layer 3 mechanism. The key idea is to mark the priority traffic so that it isnt affected by possible congestion due to the fluctuation of link usage level. The DiffServ concept is a priority marking scheme carried in the IP header.

Layer 2 QoS mechanisms are used when the traffic traverses L2-only devices (e.g., microwave radios). These devices use two 2-byte fields, tag protocol identifier (TPID), and tag control information (TCI) to mark packets with the corresponding service type.

Figure 11 displays the QoS mechanisms used in LTE network. The traffic type at the radio level is managed by the QoS class identifier (QCI). The operator defines the rules in the eNodeB by means of a QCI mapping table containing a one-to-one mapping from the QCI value to the DSCP flag value and L2 QoS codes. Then, the intermediate device uses the QoS values to treat the traffic according to their capabilities (i.e., L2 or L3 QoS mechanisms).

Figure 11 QoS mechanisms in LTE networks

Its recommended that the NaaS operator ensure that traffic is classified and marked as close as possible to its origin. In most of the cases these tasks are performed by the eNode B for the uplink traffic and by the core elements for the downlink traffic. Additionally, all the intermediate nodes must maintain the QoS markings. Its highly recommended that at least 4 different classes of traffic are defined to be used on the transport network.

4 Transport and IP Architecture Design Considerations

Analysis of multiple implementation options to be considered in the overall solution. In addition, methodologies for the most appropriate selection are provided.

4.1 Transport Technologies Selection

This section describes the architecture of different transport technologies to be used on the transport network. Moreover, it describes implementation alternatives for each technology and general guidelines to select that which is most appropriate.

4.1.1 Fiber Optic Transport Considerations

The general architecture using fiber optic as a transport technology is displayed in Figure 12. Here, the enodeB is connected to a router located at the site, commonly named as cell site gateway (CSG), which provides the transport services. The CSG is connected through a fiber link to the aggregation node located in the transport node, which sends the traffic to the mobile core. Alternatively, an L2 switch can be deployed on the cell site location, moving the CSR location to the aggregation node.

Figure 12. Transport architecture using fiber optic

For the implementation of the fiber optic architecture, three scenarios are identified for NaaS operators:

  • Fiber Construction Implies building the fiber infrastructure up to the RAN site. Thus, the design of the fiber laying must be performed.
  • Fiber Link Leasing Fiber links can be leased to a third-party with a presence in the region.
  • No Fiber Links Fiber optic may not be feasible if the cost of fiber infrastructure is out of the budget (e.g., high cost of fiber deployment in rural areas), and a third-party infrastructure cannot be used (e.g., no third-party with coverage in the area). In this case, the technology should not be considered in the later phases of network design.

In the following sections, the primary considerations for the two first scenarios are examined, and engineering guidelines are provided accordingly.

4.1.1.1 Fiber Construction

This scenario considers creating a fully dedicated, private physical network infrastructure by the NaaS operator. This approach can easily scale to handle whatever bandwidth requirements arise in the future.

On the other hand, the construction cost in rural areas is high, which may impact the overall CAPEX project. Additionally, construction time can be lengthy and requires extensive planning and project management.

The principal considerations to support the fiber infrastructure design process are described in the following subsections.

Right of Way Definition

A right of way provides the right to pass across the lands of another, usually in a strip, acquired for or devoted to building facilities such as roads, railroads, or utilities. In this sense, rights of way are closely interrelated with both civil work for fiber infrastructure and technical engineering for network planning, operation. and maintenance.

For the reasons stated above, a right of way definition must be established to prioritize the types of road (e.g., main roads, highways, electrical lines) that may be considered during later phases in the transport network design. For example, a NaaS operator can only be granted to construct fiber infrastructure in rural roads, so other types of roads (e.g., highways) must not be considered during the design phase.

Fiber Optic Maximum Allowed Distance

Fiber Optic Maximum Allowed Distance is the maximum distance permitted for a new deployment of fiber to be considered as feasible for implementation. This value must consider the technical conditions (e.g., target availability) as well as commercial considerations (e.g., maximum available CAPEX for fiber deployment).

This value is important for the overall transport network design, as it directly affects the number of feasible fiber links. The greater the value, the greater the amount of fiber infrastructure to build, which increments the cost of deployment.

Fiber Optic Interfaces

In this scenario, multiple fiber strands are available as part of the fiber infrastructure laying; this characteristic helps determine the fiber optic interface to be implemented. The fiber optic interfaces are modules used for the transmission of signals over the fiber optic medium. The complexity and cost of such modules varies according to the capacity supported (e.g., 1G, 10G, 40G. or 100G).

Additionally, a capacity safety margin must be defined to determine the capacity of exhaustion in the interface. For instance, a safety margin of 20% indicates that the interface capacity is exhausted when the real utilization hits 80% of the interface capacity. In this case, an upgrade in the interface must be considered.

Preferred Type of Fiber Laying

A new fiber deployment could require buried, aerial, or underwater cables. Moreover, cable may be in conduit, innerduct, or direct buried; aerial cables may be self-supporting or lashed to a messenger.

In rural areas, aerial cable is the most common fiber laying deployment due to its low cost and fast implementation.

Fiber Construction Summary

Table 3 provides the NaaS operator with an initial set of recommended values for the fiber construction.

Characteristic

Engineering Guideline

Right of Way Definition

New deployments must prioritize the use of main roads (e.g. highways) and consider secondary roads as subject to permissions grant.

The use of the electrical grid must be considered as priority if available.

Fiber Optic Maximum Allowed Distance

The maximum allowed distance between eNodeB and transport Node is generally 10km depending on the availability calculation and commercial agreement. For example, in a rural environment where a fiber cut occurs each 8 years per km and a MTTR of 1 day, the maximum allowed distance is 3km to maintain an availability of 99.95%.

Fiber Optic Interfaces

1G interface per Last-mile Links 10G interface per Aggregation Links *Capacity Safety Margin of 20% of interface capacity

Preferred Type of Fiber Laying

Prioritize the use of aerial fiber.

Table 3 Engineering guidelines for fiber construction

4.1.1.2 Fiber Link Leasing

NaaS operators can rely on an existing fiber deployment by leasing the fiber links to a third-party. This process implies an additional OPEX but avoids the investment in infrastructure deployment.

Here, the operation and maintenance activities of the fiber infrastructure is the responsibility of the third-party, lightening the operational workload. The monitoring to assure the service level agreement (SLA) compliance can be done over the demarcation node between the NaaS operator and the third-party network. This approach is commonly known as lit fiber leasing.

Another option for the NaaS operator is to only lease the fiber infrastructure of the third-party and install its own equipment. The approach is commonly known as dark fiber. Here, the NaaS operator can upgrade the capacity by changing the equipment at the ends of the fiber, making the link future-proof. Moreover, the fee cost remains fixed and independent of the amount of coursed traffic. Due to the nature of this scenario, some considerations of the previous section are applied.

In the following subsections, the principal considerations when leasing fiber links (lit fiber leasing) to a third-party are described.

Fiber Link Capacity

With fiber construction, the capacity of fiber link is not a constraint; several fiber strands are available and capacity exhaustion is not likely. However, the expected transmission rate on a leased fiber link must be negotiated within the SLAs.

Fiber link capacity must be appropriately forecasted to avoid surpassing the contracted limit. If the contracted capacity is exceeded, the traffic may be discarded, causing service degradation. Additionally, a capacity safety margin must be considered in the link capacity.

Furthermore, in subsequent design phases, this value determines the maximum number of aggregated sites within a single transport link depending on the traffic generated by RAN equipment.

Fiber Link Availability

The availability in a leased fiber link is another parameter to be negotiated within the SLAs. This value varies according to the network segment.

Fiber Link Leasing Summary

The typical values recommended for the NaaS operator regarding fiber link leasing are summarized in Table 4.

Characteristic

Engineering Guideline

Fiber Link Capacity

100Mbps 1Gbps per Last-mile Links 1Gbps 10Gbps per Aggregation Links *Capacity Safety Margin of 20% of link capacity

Fiber Link Availability

– 99,9% per Last-mile Links – 99,99% per Aggregation Links

Table 4 Engineering guidelines for fiber link leasing

4.1.2 Microwave Transport Considerations

The general architecture using microwave as a transport technology is displayed in Figure 13. Here, the eNodeB is connected to the radio equipment located in the site, which is connected through a MW Link to radio equipment located in the transport node. A connection to the aggregation node is reached that sends the traffic to the core elements.

Figure 13 Transport architecture using microwave

Microwave links may not be feasible due to unfavorable terrain conditions (e.g., mountainous terrain) or not suitable civil infrastructure (e.g., low rise towers). In this case, the technology should not be considered in later phases of network design.

If the NaaS operator selects microwave links as a suitable option, they should take the following considerations to support the design process.

Frequency Band Selection

There are two main categories when selecting the operational frequency band to be used in microwave links: licensed and unlicensed. These terms refer to the RF spectrum characteristics set by the respective national government regulatory body. Licensed products require regulatory approval before deployment, while unlicensed products can be deployed without any such approval.

A comparison between using licensed and unlicensed frequency bands is shown in Table 5.

Characteristic

Unlicensed Bands

Licensed Bands

Licensing Cost

No licensing cost

License cost, in some markets spectrum charges can be high.

Total Time to Deploy

Very low as no license needs to be applied for and be granted.

High, as license needs to be applied and granted.

Interference Level

Medium to High due to external systems using the same frequency, causing link performance and throughput to be compromised.

Very low interference and no associated performance degradation.

Link Capacity

Low-Medium (~100-200 Mbps) due to limited available spectrum (typically 40MHz, but some equipment support up to 80MHz with some sensitivity constraints)

High (~1Gbps) due to larger communication channels (up to 112MHz) and reduced interference

Latency

High (~2-10 ms) due to time-division duplex (TDD) nature of most of unlicensed link (link cannot transmit and receive simultaneously)

Low (~0.1ms) due to use of frequency division duplex (FDD) links

Equipment Cost

Generally low due to mass-market WiFi chipsets

Higher, as specialized chips are required.

Table 5 Comparison of licensed and unlicensed frequency bands in MW

The final selection of the frequency bands to be considered in later design phases depends on the available budget and general capacity requirements. However, its highly recommended to select a small number of frequency bands (e.g., twoone for small distance links and one for large distance links) to reduce complexity in the design phases.

Licensed and unlicensed microwave technologies are not mutually exclusive. This means the NaaS operator can decide to use unlicensed bands for links having small distance and low capacity requirements and licensed bands for long-range links with high capacity requirements.

Microwave Link Capacity

Microwave link capacity is the expected transmission rate on a microwave link that must be defined for each network segment (last-mile and aggregation).

Detailed on the Tx LLD module, a link design must be performed to obtain the specific values (e.g., modulation, Tx power) to reach the expected capacity. Additionally, a capacity safety margin must be defined to determine the capacity of exhaustion in a microwave link.

Permissible Topologies

Figure 14 displays physical topologies that can be supported by using microwave links.

Figure 14 Microwave topologies

The maximum number of sites that can be aggregated in a topology depends on the following characteristics:

  • Aggregated traffic on the link The higher the capacity per eNB, the fewer eNBs can be aggregated. In addition, depending on the aggregated traffic, the last hop to the transport network might require deployment of an aggregation link.
  • Target availability The higher the availability target, the fewer number of hops that can be deployed in the topology. One alternative to increase availability is to implement the ring topology, as it provides path redundancy (however, the number of links may increase).

The methodologies to validate the feasibility of disparate physical topologies are included within the Tx HLD Module.

Microwave Transport Summary

Table 6 summarizes typical values recommended for the NaaS operator regarding microwave link deployments.

Characteristic

Engineering Guideline

Frequency Band Selection

– Unlicensed 5GHz for distances < 5km – Licensed 15GHz for distances in the range of 5km-15km – Licensed 7GHz for distances in the range of 10km-25km

Microwave Link Capacity

– 50Mbps – 100Mbps per Last-mile Links – 100Mbps – 250Mbps per Aggregation Links *Capacity Safety Margin of 20% of link capacity

Allowed Topology

– Star, chain and tree topologies. *The feasibility of each topology must be validated during HLD Design.

Table 6 Engineering guidelines for microwave link deployment

4.1.3 Satellite Transport Architecture

Architecture using satellite as a transport technology is shown on Figure 15. Here, the eNodeB is connected to a very small aperture terminal (VSAT) located at the site, which is connected through a satellite link to the terrestrial satellite hub. From that point a connection to the aggregation node is reached that in turn sends the traffic to the core elements.

Figure 15 Transport architecture using satellite

The most common implementation for satellite technology is to use a satellite service provider. Principal considerations when selecting and negotiating with such a provider are described in the following subsections.

Type of Satellite Orbit

Table 7 presents a comparison among different types of satellite orbits.

Orbit

GEO

MEO

LEO

Distance

35,800 km

2,000km-35,000km

160km-2,000km

Latency

250-500 ms

60-250 ms

30-50 ms

Throughput

Up to 500 Mbps

Up to 800 Mbps

1Gbps+ *(See NOTE)

Handover Periodicity

No handover

2-3 hours

15-20 min

Table 7 Satellite orbit comparison

*As LEO systems have not been launched commercially, the 1Gbps value is a theoretical reference only.

Despite latency constraints, GEO orbits are preferable as they provide larger footprints and no satellite tracking mechanisms or handover are required. Additionally, LEO systems have not been launched commercially but may become an attractive solution when available.

Satellite Frequency Band

The satellite frequency band determines the available bandwidth and potential throughput in the satellite link. Ka-band (~30GHz uplink and ~20GHz downlink) and Ku-band (~14GHz uplink and ~12GHz downlink) are the most-used bands for mobile transport.

Most Ka-band satellites offer 10x the throughput available on traditional Ku-band satellites. The offered throughput increment also decreases the fee price of the link. Moreover, Ka-band antennas have higher gain, resulting in less expensive VSAT equipment.

On the other hand, the higher frequency, the more a signal is susceptible to rain fade. So adverse weather conditions impact the Ka-band much more than at lower frequencies

For the reasons stated above, Ka-band is preferable for implementation as the frequency band for satellite transport technology. However, Ku-band can be the only feasible option in extreme meteorological territories and where there is a lack of Ka-band coverage.

Satellite Link Capacity

Satellite link capacity is the expected link transmission rate that must be defined. This value should be used as reference in negotiating SLAs with the provided satellite service.

Additionally, a capacity safety margin must be defined to determine the capacity of exhaustion in a satellite link. For instance, a safety margin of 20% indicates that link capacity is exhausted when real utilization hits 80% of the link capacity.

Satellite Transport Summary

The typical values defined for satellite link leasing scenario are summarized in Table 8.

Characteristic

Engineering Guideline

Type of satellite orbit

Preference on GEO Satellites

Satellite Band

Ka Band is preferred

Satellite Link Capacity

– 5Mbps – 15Mbps Downlink – 1Mbps – 5Mbps Uplink

Table 8 Engineering guidelines for satellite link leasing

4.2 Transport Service Providers Evaluation and Selection

A NaaS operator can elect to rely on a third-party network when there is a well-established transport service providing expected transport services along with the corresponding SLA. This section focuses on technical aspects to evaluate transport service offers. The complete process to include a transport service provider is included in the Procurement module.

There are two possible scenarios to include a third-party network:

  • Last-mile link implementation This case only applies when the third-party provides the infrastructure from the RAN site up to the core element.
  • Aggregation link implementation This is the most common scenario, where the NaaS operator deploys the last-mile link up to a third-party aggregation node. Then its the third-partys responsibility to transport the traffic to the core elements.

From a technical perspective, Table 9 shows the typical requirements a transport service provider must satisfy that must be defined according to required network service (e.g., last-mile or aggregation link implementation scenario).

Evaluated Characteristics

Availability

Reliability and uptime

Time-to-Repair

Time-to-Deploy

Capacity

Guarantee Capacity

Latency

Guarantee Latency

Security

Security mechanism

Table 9 Typical characteristics to be evaluated in a transport service provider

The decision to include an external transport service provider is also affected by project financial constraints. By leasing an external network, considerable investment in infrastructure might be avoided (e.g., CAPEX) and the operation and maintenance activities become responsibility of the third-party, thereby lightening the operational workload. However, an additional OPEX must be considered along the lifespan of the network.

Additionally, the NaaS operator can use the Transport Service Providers Evaluation Matrix Template to support technical evaluation of disparate transport service providers.

4.3 IP Networking Design

Figure 16 illustrates the specified 3GPP connections between eNodeB and core elements that are carried over the transport network and the specified protocol stack.

Figure 16 3GPP logical interfaces protocol stack

To support the Figure 16 scenario, the NaaS operator must define the protocols and technologies to be used at the various layers. The following subsections present available options to be considered along the layers when considering IPv4 protocol. Additional IPv6 protocol considerations are provided at the end of the section.

4.3.1 L1 / Physical Interfaces

Using Ethernet ports as physical interfaces reduces the need to install or upgrade physical ports, as IEEE standards include data rates for 1Gbps (gigabit Ethernet), 10Gbps (10G), 40Gbps (40G), and 100Gbps (100G). For instance, an initial 1Gbps port is adequate and can support growth in the first years of the network.

In addition, Ethernet supports both electrical (twisted-pair cable) and optical fiber interfaces for physical connection.

The Ethernet physical port and Ethernet L2 frame are basically agnostic to the higher layer protocol (i.e., protocol type depends on the upper layer protocol). That means either IPv4 or IPv6 can be carried.

4.3.2 L2 / Data Link Layer

As the majority of LTE mobile equipment uses Ethernet interfaces for transport, Ethernet-based services are most suitable to be implemented in the transport network.

4.3.2.1 Virtual LAN (VLAN) Support

Virtual LANs are used to separate traffic logically, or based on functionalities, over the same Ethernet port (e.g., logically separate O&M from the user/control traffic into a different VLAN). This is achieved by adding a VLAN tag to each Ethernet frame so that customer VLANs can be separated from those of the provider. This tag is used by other layer 2 devices to switch traffic in the network.

Including a VLAN tag increases the Ethernet frame header by two bytes, which needs to be taken into account when calculating maximum transfer unit (MTU) at the Ethernet layer.

Most base stations support a flexible way to bind eNB applications (S1/X2 U-plane, S1/X2 C-plane, M-plane, S-plane) arbitrarily to either eNB interface address(es) or eNB virtual (loopback) address(es).

This flexibility lets base stations be configured according to the transport services offered by the transport network, but also applies traffic separation (e.g., M-plane from U/C-plane traffic) as required. eNB interface IP address(es) can be assigned either to one or more physical interface(s), or one or more logical interface(s).

A physical interface is typically provided by an Ethernet port, whereas a logical interface is provided by a VLAN termination. Disparate interfaces as well as VLANs belong to different IP subnets.

Many configurations are possible, but for sake of simplicity only two exemplary configurations are shown in Figure 17:

Figure 17 Example IP configurations of eNB

In the left-hand example, one IP address terminates control, user, and synchronization plane traffic; a second one does solely for M-plane traffic. As both IP addresses are running over different logical interfaces created by two VLANs, the addresses have to belong to different IP subnets. In the right-hand example, separate logical interfaces (VLAN) and IP addresses are used for each plane (user, control, management and synchronization).

The recommended configuration is to maintain separate logical interfaces and IP addresses for each plane. In this case, each will be mapped to a unique IP address in the network. Table 10 shows a typical configuration of VLAN tags used to identify the traffic.

Plane

VLAN Tag

Control Plane (CP)

101

User Plane (UP)

102

Management (OAM)

103

Synchronization (SYN)

104

Table 10 Typical VLAN configuration

4.3.3 L3 / Routing Layer

LTE specifications define an IP-based protocol stack for the logical interfaces (S1 and X2). Both versions of IP protocol (IPv4 and IPv6) are included in 3GPP standards. But throughout this text IP mentions refer to IPv4 unless specifically mentioned.

Elements in the network need the next hop (router or interface) toward a given destination. This information is learned dynamically with the help of a routing protocol. However, 3GPP has no guidance for the use of routing protocols. An IP host (e.g., eNodeB) might have a static, preconfigured entry only for the default gateway, or it might run a routing protocol. The following subsections examine approaches to implement the network routing mechanism.

Static Routing

The simplest method is the use of static routes. Their main advantage is that theyre persistent regardless of any network perturbation. The required state is minimal and the absence of a control plane makes the device more scalable. However, persistent state does not adapt to network topology changes.

A static route is recommended for endpoints where a single connection is available. For instance, in a scenario where only one cell site router is deployed, the eNodeB can only send the traffic to this element. Once a second interface is added, a dynamic means of managing the state of the device routing table is required.

Routing Protocols

A routing protocol learns routes to destination networks from other routers in the network. This permits the elements to dynamically adapt to varying network conditions (e.g., link failure) and is more scalable as the network grows. One common classification of routing protocols is based on protocol role in the network:

  • Interior Gateway Protocols (IGP) Used to exchange routing information within an autonomous system (AS). The open shortest path first (OSPF) protocol is a link-state routing protocol that uses a hierarchical routing modelwhere a centralized routing domain is defined within an autonomous system. OSPF can be used in the aggregation segment when more complex topologies are presented.
  • Exterior Gateway Protocol (EGP) Used to exchange routing information between AS. The border gateway protocol (BGP) is a distance vector used to advertise the route prefixes outside the organizational boundary, without revealing the internal routing characteristics. BGP is commonly used to exchange prefixes assigned to mobile subscribers to the providers internet peer.

Supporting Protocols

Additional supplementary protocols included in the IP stack must be considered during implementation, such as:

  • Internet Control Message Protocol (ICMP) ICMP is an integral part of the IP protocol and all routers must support it. With ICMP, hosts and routers can report errors and exchange diagnostic, control, and status information.
  • Address Resolution Protocol (ARP) ARP is an L2 Ethernet broadcast that asks for the MAC address corresponding with the next-hop IP address.

Figure 18 depicts a typical implementation for L3 routing protocols.

Figure 18 Typical l3 routing protocol implementation

4.3.4 L4 / Transport Layer

The most commonly used transport layer protocols are examined in the following subsections.

User Datagram Protocol (UDP)

With the mobile transport application, UDP is the protocol on top of IP in user plane, and interfaces with the radio layer protocols (GTP-U protocol on S1 and X2). UDP makes no assertion about the reliable delivery of data.

Stream Control Transmission Protocol (SCTP)

SCTP (stream control transmission protocol) provides reliable transport service for messages over the IP network. In the mobile transport network, its used for signaling traffic. LTE signaling, S1 and X2 control plane, is over SCTP.

Transmission Control Protocol (TCP)

TCP isnt included in the mobile transport protocols stacks, since the user plane is over UDP/IP and control plane over SCTP/IP. However, TCP can still be used in the M-plane.

TCP is also used as a reliable transport mechanism for some of the IP and MPLS control plane protocols. For example, BGP uses TCP.

4.3.5 IP Protocol Stack Summary

Table 11 summarizes recommended protocols for consideration within the reference architecture.

Description

Selected Protocol

L1 / Physical Interfaces

Ethernet Interfaces in all ports: – 1G for last-mile interfaces. – 10G for aggregation interfaces.

L2 / Data Link Layer

– Ethernet – Support of 802.1q and 802.ad used in VLAN.

VLAN Definition

Four VLANs, one for each of the following planes: – User plane – Control Plane – Management Plane – Synchronization Plane

L3 / Routing Layer

– Support of IPv4 and IPv6 dual stack – Static routing for a single site in the last-mile. – OSPF in aggregation network – BGP for connection to 3rd Party Networks – Support for complementary protocols (ICMP, ARP)

L4 / Transport Layer

– UDP – SCTP – TCP

Table 11 Recommended IP protocol stack in mobile operator

Additionally, the NaaS operator can use the Wizard for Tx & IP Architecture to support them in defining required protocols to be implemented.

4.3.6 High Level IP Plan Design

An IP address distribution plan must be defined to simplify overall network IP address management, support several technologies, and be able to perform efficient troubleshooting. An IP address distribution plan is a document developed by the NaaS operator that displays how the available IP address universe will be distributed in a way that supports the required services. This plan should satisfy the following conditions:

  • It should be simple and follow logical rules.
  • It should be scalable up to the forecasted number of eNodeBs and network elements to be deployed considering an additional margin for future expansions.
  • It should enable simple routing configurations and efficient route summarization.

The following subsections illustrate methodology to generate such a plan using IPv4. A similar process can be performed when using IPv6 (considering that 16 octets are available).The specifics of this process are in the Primer on IP Planning Principles.

Network Range Definition

The first step to generate the overall IP address distribution plan is to define the network range to be used in allocating IP addresses for each element in the NaaS operator network.

The Internet Assigned Numbers Authority (IANA) reserves a number of IPv4 network ranges as private. Theyre reserved for organizations wanting to build an internal network and are:

  • 10.0.0.0 10.255.255.255 (10.0.0.0/8)
  • 172.16.0.0 172.31.255.255 (172.16.0.0/12)
  • 192.168.0.0 192.168.255.255 (192.168.0.0/16)

The NaaS operator can select any of these ranges to be used within their own network. However, its highly recommended to use the 10.0.0.0/8, which permits the greatest flexibility.

IP Subnetting

Subnetting is used to partition a network address space into smaller segments (e.g., divide one segment /8 into 16 segments /12). In this way, the NaaS operator can separate the network space addresses for each network element type (e.g., eNodeBs, transport devices, core elements).

There are two primary conditions to consider in creating the IP address distribution plan the type of network elements and their geographic distribution. In this sense, there are two main approaches:

  • Location-First In this approach, contiguous subnets are allocated to each of an operators defined regions, after which the subnet for a given region is divided among the network element types. The most significant bits of the second octet are used to identify the geographical zone.
  • Element-First Here, each network element type is allocated with one subnet divided among the regions. The most significant bits of the second octet are used to identify the network element type.

An efficient application route summarization can be done using the element-first approach, so its recommended for creating an IP distribution plan supported by the NaaS operator.

Figure 19 shows a typical IP distribution plan for a mobile operator. It considers four types of elements and assumes the operator has divided its network in two geographic zones (north and south).

Figure 19 Typical IP distribution plan in mobile operator

Using the Figure 19 guide, Table 12 shows the IP distribution in decimal notation.

Type of element

Zone

Range

eNodeB

North

10.0.0.0/12

South

10.16.0.0/12

Router Loopback

North

10.32.0.0/12

South

10.48.0.0/12

MW_O&M

North

10.64.0.0/12

South

10.80.0.0/12

Core

North

10.96.0.0/12

South

10.112.0.0/12

Reserved

10.128.0.0/9

Table 12 Typical IP distribution plan in mobile operator

Additionally, a NaaS operator can use the High Level IP Distribution Plan Widget in the creation of the IP distribution plan.

4.3.7 IPv6 Considerations

The subsequent sections analyze the principal considerations regarding IPv6 protocol.

IP Addressing in Transport Network

A main IPv6 driver is the unavailability of public IPv4 addresses. The latter are a scarce resource, and several layers of network address translation (NAT) could be needed. This complicates the network and also introduces issues for many applications.

However, the IP addresses used for connectivity within the NaaS operator network dont need to be public (apart from some interfaces in the mobile core), and thus there is no pressing need for a larger address space (in general). Thus, network elements within the transport network can use IPv4 addresses.

IPv6 Protocols in Transport Network

Even though both IPv4 and IPv6 are IP protocols, they differ. IPv4-only devices arent capable of routing IPv6 packets. Network elements must support dual-stack configuration to support both protocols. Nodes that implement dual-stack maintain two protocol stacks (IPv4 and IPv6) working in parallel, thereby enabling the end of the service or router to operate through either protocol.

Furthermore, separate versions of routing protocols must be maintained for each IP version (e.g., Open Shortest Path First version 3 (OSPFv3)).

Support for IPv6 Native Applications

In addition to the IP layer present in the transport network, end user applications use IP. The user IP layer is transparent to the transport network, as it is encapsulated within the radio network layer protocols. User IP layer can be IPv6 even though the transport network is using IPv4. In other words, the two layers (end user IP and transport network layer IP) are isolated.

Figure 20 shows how IPv4 and IPv6 user applications can use the transport network independently of the implemented IP version. Here the devices required to support both IP protocol version (i.e., dual-stack devices) are UE, eNodeB and mobile core elements.

Figure 20 Encapsulated user traffic over transport network

In general, its recommended to implement dual-stack network elements to support legacy services and applications. This will facilitate migration to a full IPv6 network.

4.4 Redundancy and Fault Detection Considerations

The present section includes examining diverse redundancy mechanisms, both physical and logical, as well as fault detection methods to evaluate their applicability in the transport network.

Microwave Protection Scheme

The basic microwave protection scheme is the 1+1, which translates into radio redundancy, where two radio units (RUs) create a single radio link. In the event of RU failure, the hot standby radio maintains the site service. The reason for implementing such redundancy is because recovery time can be lengthy, especially in rural areas. MTTR must include site travel time (with the correct spare parts) and the time it takes a qualified technician to climb and perform the replacement at height.

On the other hand, the main drawback of this scheme is that two RUs must be deployed for each microwave link; this elevates total CAPEX of the deployment. For this reason, 1+1 protection should only be considered for critical links (e.g., MW links in the transport aggregation segment).

Ethernet Link Aggregation

Link aggregation combines multiple Ethernet links into a group, which is seen as a single link. The benefits are an increment in capacity as well as resiliency, as failure of a single link is tolerated. Link aggregation also permits load sharing that is otherwise not supported by Ethernet.

Link aggregation group (LAG) can be implemented in aggregation segments as a means of redundancy to reach the target availability.

Router Redundancy Protocols

Router redundancy protocols (e.g., virtual router redundancy protocol (VRRP) or hot standby router protocol (HSRP)) can be considered a form of L3 redundancy. The operational principle is to let multiple routers share a virtual IP and MAC address. One router is elected as the master, which assumes the role of forwarding packets sent to the virtual IP address. The remaining routers act as backup. If the master fails, one of the backups becomes the master.

Router redundancy protocols can be implemented in aggregation segments as a redundancy method to reach the target availability. The specific selection of the protocol varies according to the selected Tx equipment.

Bidirectional Forwarding Detection (BFD)

Bidirectional forwarding detection (BFD) is a fault detection protocol providing a low-overhead, short-duration method to assess a communication failure between devices and notify upper-layer applications (also called clients). BFD is a detection protocol that can be enabled at the interface and routing protocol levels.

For instance, a network running OSPF takes several tens of seconds to recover from a failure, using the default parameter settings. The main delay component is the time to assess a failure using timeout-based detection. In contrast, BDF typically detects a link failure in 50ms, which leads to a significant decrease in network recovery time.

Its recommended to enable BFD in interfaces and routing protocol levels only on those elements located in the aggregation segment.

Table 13 summarizes recommendations to be considered within the reference architecture.

Description

Selected mechanism

Redundancy mechanisms

Microwave Protection Scheme 1+1 in aggregation links

Ethernet Link Aggregation in aggregation links

Router Redundancy Protocols in aggregation nodes

Fault detection mechanisms

Bidirectional Forwarding Detection in aggregation nodes

Table 13 Engineering guidelines for redundancy and fault detection

Additionally, the NaaS operator can use the Wizard for Tx & IP Architecture in determining what to implement on their network.

4.5 QoS Mapping and Scheduling Considerations

This section presents a methodology to establish general rules for mapping between RAN QCIs and L3 QoS markings. Additionally, it examines packet scheduling methods to determine their implementation suitability.

4.5.1 QoS Mapping Function

The NaaS operator controls the quality of service by defining QoS class identifiers (QCIs) according to characteristics of the service (section 3.3). In the transport network, a mapping among QCIs to DSCP bits and between DSCP bits and L2 p-bits are performed as illustrated in Figure 21. The ability to map QoS codes between layers is essential to ensure interoperability and providing E2E QoS.

Figure 21 Mapping function of QoS values among network layers.

Per-hop behavior (PHB) is an important concept to define the forwarding behavior applied to the traffic types. A node performs the same PHB for packets having the same DSCP value. A NaaS operator implementation can use the following PHB types:

  • Expedited Forwarding (EF) EF traffic has the highest priority above all other traffic classes. The EF PHB applies to services that require a short delay, low jitter, and low packet loss rate, such as voice and signaling traffic.
  • Assured Forwarding (AF) Assured forwarding provides assurance of delivery as long as the traffic doesnt exceed some subscribed rate. Traffic that exceeds the subscription rate faces a higher probability of being dropped if congestion occurs.
  • Best-Effort (BE) The BE PHB focuses only on whether packets reach the destination, regardless of transmission performance.

Table 14 lists the recommended QoS code mapping function among different layers. It includes the per-hop behavior value that determines treatment for each traffic type.

Type of Traffic

QCI

p-Bit

DSCP

PHB

Data

9

0

0

BE

Voice

1

5

46

EF

eNB-Control

X

4

34

AF41

eNB Synchronism

X

5

46

EF

Management

X

3

26

AF31

Table 14 Recommended mapping function among QoS code

4.5.2 Packet Scheduling Mechanisms

Every time packets enter into devices faster than they can exit, there is a possibility of congestion. Packet scheduling handles congestion management within the transport network and must be implemented in any interface that can experience congestion. When congestion occurs, packets should be temporarily stored or queued in temporary storage for subsequent scheduling, as shown in Figure 22.

Figure 22 Packet scheduling into different queues

Commonly used queue types priority queuing (PQ), that serves packets in strict priority order, and weighted fair queuing (WFQ) that divides the interface bandwidth among the disparate flows according to their priority, ensuring a fair bandwidth distribution for all applications.

However, buffering memory is a limited resource on any interface. By queuing, buffers fill up and packets can be discarded as they arrivecalled tail dropor packets can be selectively dropped before all buffers are filled, known as weighted random early detection (WRED).

Table 15 shows recommended congestion management configuration for traffic types in the transport network.

PHB

Assign Queue

Packet drop mechanism

EF

Priority queuing (PQ)

Tail drop

AF4

WFQ

WRED

AF3

WFQ

WRED

AF2

WFQ

WRED

AF1

WFQ

WRED

BE

Default Queue

RED

Table 15 Recommended configuration for congestion management

Additionally, the NaaS operator can use the QoS Mapping and Scheduling Mechanisms Template to define required implementations.

4.6 Additional Architectural Design Considerations

This section provides guidance on additional design considerations that must be considered to support the overall transport and IP network solution.

4.6.1 Synchronization

Synchronization is fundamental in LTE network operation as its failure can result in spectral inefficiency and service degradation. Phase/time and frequency are synchronization types for a mobile network.

FDD-LTE networks without special radio features can be deployed with frequency-only synchronization. Here the synchronization signal is needed only for keeping the carrier frequency within required limits.

TDD-LTE and certain advanced LTE features (e.g., coordinated multi-point (CoMP)), require that all eNBs are phase-synchronized; they must start radio frame transmission at exactly the same instant.

Global Navigation Satellite System (GNSS)

In satellite navigation systems (e.g., GPS and GLONASS), a GNSS receiver derives frequency and calculates time from the satellite signals. Synchronization equipment then uses it as a reference for network timing.

The GNSS receiver can be a standalone device or embedded in the base station. A GNSS receiver can also be integrated into a collocated or nearby cell site router (CSR).

GNSS implementation requires that the receiver location provides an unobstructed view of the corresponding GNSS satellites. The cost to deploy a GNSS receiver in each node could impact overall CAPEX deployment. For this reason, this synchronization method can be considered when a satellite link is used as transport technology.

Precision Time Protocol (PTP)

The IEEE 1588 precision time protocol (PTP) was developed to enable accurate distribution of time and frequency over packet-based networks. PTP requires the implementation of a server (PTP grandmaster clock) as the primary reference source for all PTP clients.

For sites not having a GPS receiver, its recommended to adopt PTP as a clock synchronization protocol. Its a standardized protocol, is accurate, and offers lower cost. The clock signal is transmitted transparently through the transport network.

Synchronization Summary

Synchronization selection represents a trade-off among cost and complexity. For a small network deployment it might be easier to implement a GNSS receiver in all base stations and avoid the PTP infrastructure and configuration.

Table 16 provides the recommended synchronization methods to be implemented in a transport network.

Description

Selected mechanism

Synchronization mechanisms

Global Navigation Satellite System (GNSS) preferred when satellite transport and unlicensed backhaul is used

Precision Time Protocol (PTP) for terrestrial transport

Table 16 Recommended synchronization mechanisms in the transport network

4.6.2 MTU Considerations

The maximum transmission unit (MTU) is the largest size packet that can be processed by nodes within the transport network. When a packet to be transported exceeds the MTU size of the egress interface, it needs to be fragmented so as to be sent forward. But packets can be dropped when MTU configuration doesnt match on the intermediate nodes, thereby generating a service degradation.

Because connectivity between the eNodeB and core elements uses GTP-U tunneling, the packets grow when sent over the transport network. Additional means are required to provide the LTE service (e.g., VLAN tagging, IPSec); increasing total packet size and fragmentation cannot be completely avoided. Thus, matching the path MTU with the IP packet sizetaking into account the additional headers (e.g., GTP, VLAN, IPSec headers)can reduce the possibility of fragmentation.

In Ethernet networks, the default MTU value is 1500 bytes. In some cases, using protocols that increase the frame header (e.g., VLAN, IPSec, mini-jumbo (3000 bytes), or jumbo frames (9000 bytes)) might be required to prevent fragmentation.

The NaaS operator should evaluate the MTU to be used along the transport network, considering all additional headers resulting from architectural decisions. Furthermore, when the traffic traverses a third-party network, its important to verify that the defined MTU value is set across every link in the external network to prevent traffic being dropped.

4.6.3 Security

This section evaluates transport network security options to determine selection guidelines in achieving the most suitable implementation.

Network Element Security

Its foundational for a network to implement security at the network element level. This requirement applies to any network-connected node.

Network element security covers multiple aspects of node functionality. Some require attention by equipment manufacturers at an early design and implementation phase, while others need to be considered by a NaaS operator during deployment and maintenance phases.

The following list includes the most relevant actions that should be performed as part of deployment and maintenance of a network element:

  • Support of user authentication and controlled access levels.
  • Support secure access by the management system user (e.g Secure Shell (SSH), Simple Network Management Protocol (SNMPv3))
  • Unnecessary physical interfaces are disabled.
  • Unnecessary services and ports are disabled.
  • Support of CPU overload control.
  • Default passwords are changed.
  • Passwords are regularly changed and only strong passwords are used.

Additional actions might be recommended by the transport equipment OEM.

IPSec Framework

IPSec is not a single protocol, but rather a collection of protocols and services that provide security for IP networks. It includes protocols such as authentication header (AH) and encapsulating security payload (ESP), internet key exchange (IKE), and certain algorithms for authentication and encryption.

IPSec provides secure services for IP packets mainly through encryption and authentication. These include user data encryption, data integrity authentication, data origin authentication, and anti-replay.

To provide these services, the functionality of a security gateway (SEG) is required. This entity could be a dedicated network element or be included in other elements depending on the specific implementation. The typical deployment considers the SEG in front of the evolved packet core (EPC). Figure 23 shows the architecture of the transport network implementing a dedicated SEG.

Figure 23 Transport architecture implementing IPSec framework

The IPSec framework can be implemented in the following scenarios:

IPSec framework can be implemented in the following scenarios:

  • IPsec for protecting the LTE control traffic (S1-C, X2-C) This scenario should be considered by NaaS operators with self-built transport networks or with eNBs in areas secure enough to just encrypt the signaling traffic.
  • Full IPsec protection (control and user traffics) This scenario should be considered when the transport services are leased from a 3rd Party or when traffic traverses by untrusted networks The main drawback of this scenario is that the impact of IPsec is significant on the U-plane, as additional overhead of about 14% is introduced. Furthermore, IPSec requires implementing more powerful hardware capable of encrypting and decrypting traffic in real-time. This can increase network equipment cost.

Security of Management and Synchronization Plane

A secure mechanism to protect the M-plane (i.e., OAM traffic) must be available even if the transport network scenario is considered trusted by the NaaS operator. The M-plane can be protected by transport layer security/TLS mechanisms.

On the S-plane side, the following recommendations must be considered:

  • If handling synchronization traffic in a dedicated VLAN isnt supported by the Tx equipment, this traffic can be transported together with the control flows.
  • If synchronization is provided with PTP technologies, some 1588v2 built-in security mechanisms can be used.
  • Its not recommended to transport PTP packets in an IPsec tunnel, as this has some impact on the delay variation.

Recommended security mechanisms for implementation in the transport network are shown in Table 17.

Description

Selected mechanism

Network Element Security

Mandatory Network Element Security Mechanisms in all nodes in the network.

User and Control Plane Security

Full IPsec protection when an untrusted 3rd party is used.

Management Plane Security

TLS protocol required for management plane

Synchronization Plane Security

PTP inbuilt security mechanisms with no encryption for synchronization plane messages

Table 17 Recommended security mechanisms in the transport network

5 Transport and IP Architecture Specification

Methodology to generate a final report that consolidates outputs from this module:

  • Transport & IP Network Reference Architecture.
  • Transport & IP Network Requirements Specification.

The corresponding Templates are included as part of the methods of engagement.

5.1 Transport and IP Network Reference Architecture

The main deliverable is the reference architecture that contains the overall technical solution, engineering guidelines, technologies, and concepts required to implement the Tx transport network. The reference architecture must contain the following aspects:

  • Overview High-level description that summarizes the Tx network architecture.
  • Transport Technologies Selection A description of selected transport technologies to be considered, as well as engineering guidelines to be used during design phases.
  • IP Networking Design A view of protocol stacks to be implemented in the various transport network segments.
  • Redundancy and Fault Detection MechanismsDescription of the redundancy and fault detection mechanism set to be considered in the network implementation.
  • QoS Mapping and Scheduling Mechanisms Description of the QoS mapping and scheduling mechanism set to be considered in the network implementation.
  • Additional Architectural Considerations Additional architectural considerations (e.g., security and synchronization) to be considered in the network implementation.

5.2 Reference Architecture Generation


1. Core Architecture Introduction

The Transport & IP Architecture Module aims to provide a frame of reference and enable the NaaS Operator to define the overall solution of the network that interconnects the Radio Access Network (RAN) towards the Mobile Core and data centers, including the technical aspects of the IP Networking solution.

This module provides an overview of the Transport & IP Network fundamentals focusing on the required functional and operational aspects to support the LTE service. Moreover, it provides guidelines to translate these requirements into a Transport & IP Architecture that better suits the NaaS operator capabilities and constraints.

The output of this module is the Transport & IP Reference Architecture that includes the overall solution description as well as a formal Requirements Specification to guide subsequent design and vendor selection phases.

1.1 Module Objectives

This module will enable the NaaS Operator to define the reference architecture of the mobile Transport & IP Network. The specific objectives of this module are to:

  1. Provide an information base and fundamentals of the architectural elements of the Transport & IP Network.
  2. Guide the NaaS operator to determine the Transport & IP Architecture functional and operational requirements to provide an LTE Service.
  3. 3. Provide instruction to define and specify the most appropriate Transport & IP Network Architecture according to NaaS Operators requirements, capabilities and business model.
  4. 4. Provide guidelines to generate a consolidated Reference Architecture and Requirements Specification for evaluation and vendor selection for the Transport & IP Network.

1.2 Module Framework

The NaaS PlayBook Framework in Figure 1 describes the structure, interactions and dependencies among different NaaS operator areas.

The Strategic Plan & Scope provides business context for the NaaS Operator and serves as a guideline for the definition of the High-Level Network Architecture, which in turn drives the strategic decisions to forthcoming phases.

The Transport & IP Network Architecture Module is included within High Level Architecture and has direct relation with RAN and Mobile Core Architecture as well as Network Monitoring Architecture. The generated output of this module will serve as a required input for the Network Design Modules.

Figure 1. Module Framework

The Transport & IP Network Architecture functional view is shown in Figure 2, where the main components are depicted, as well as the inputs that drive the definitions made throughout the development of the Transport & IP reference architecture.

Figure 2. Transport & IP Architecture Framework

The model structure is divided in four sections. Section 1 provides an overview and fundamentals involved in the design of the Transport & IP Architecture. Section 2 focuses on the functional and operational requirements of the transport network. Section 3 discusses the possible variations to implement the Transport & IP network and methodologies to perform the selection that best suits NaaS operator conditions. Finally, Section 4 illustrates how to integrate previous elements into a formal Reference Architecture and Requirements Specification.


2 Transport & IP Network Architecture

General overview of Transport and IP Network Architecture and main transport technologies to be implemented in the network. In addition, the particularities of the IP Network are examined.

2.1 Transport Network Overview

In a mobile environment, the Transport Network interconnects different networks including the RAN (Radio Access Network), data centers and external networks. Figure 3 displays the architecture of a typical transport network.

Figure 3. Typical Mobile Network Architecture

Mobile networks are ubiquitous and support a mix of different types of traffic originating from and terminating to mobile devices. All this traffic must be conveyed between the mobile cellular base stations through the transport network up to the Mobile Core.

For this reason, different segments exist on the transport network which for most of NaaS Operators can be classified as: last-mile and aggregation level. Additionally, it may exist a connection provided by a 3rd party network, which can provide these last mile or aggregation. The network segments are mostly defined by the technology and topology used within it. The key network segments characteristics are:

  • The last-mile and aggregation segments provide the connectivity to the eNodeB at the cell sites and are predominantly based on three and chain topologies built with microwave radios and optic fiber links.
  • Optionally, a 3rd party network can be used to transport traffic to the Mobile Core. A typical scenario to implement this scenario presented when the NaaS operator does not have network infrastructure in the geographical zone. In nearly all cases, the 3rd party network is an IP/MPLS network.

Figure 4 displays the structure of the transport network:

Figure 4. Basic structure of a mobile transport network

The structure displayed in Figure 4 considers different physical technologies and topologies. Moving left to right, the last-mile connects a demarcation device, usually deployed at the cell site, to a first stage of traffic grooming and concentration. The aggregation network, in turn, further aggregates traffic, adapts any technology change and provides the hand-over point to a higher level of aggregation network.

The term physical connectivity is used to represent that any technology can be used to connect nodes; common physical connectivity includes fiber, microwave, and satellite links. In addition, on top of the physical layer a networking layer is implemented which embraces all the possible logical architectures needed to steer LTE traffic and applications.

2.1.1 LTE Flattened Transport Architecture

Traditional cellular networks (e.g. 2G and 3G) are hierarchical in nature, that is, there is no direct communication between the base station and the core elements. Base stations communicate with their respective radio controllers which are in turn, connected to the core elements.

Long-Term Evolution (LTE) defines a flat network architecture that eliminates the radio controllers. Thus, controller functions are redistributed to the core elements and the base stations (eNodeB). For this, two logical S1 interfaces between the eNodeB and the evolved packet core (EPC) are implemented: the S1-MME to the Mobility Management Entity (MME) and the S1-U to the Serving Gateway (S-GW). MMEs handle authentication and signaling aspects of the mobility management control plane. S-GWs terminate the user plane traffic. Figure 5 illustrates the LTE network architecture:

Figure 5. LTE Flattened Transport Architecture

Figure 5 illustrates an overview of the logical planes which are needed in the transport network. It can be seen that different kinds of user and control planes are transported across the transport network. In addition, the management traffic which consists of the data needed for FCAPS (fault, configuration, accounting, performance, security) support of the network elements must be considered. Finally, when synchronization is arranged by the transport network (e.g. Precision Time Protocol (PTP)), these packets need to be delivered to the eNB from the timing server.

2.2 IP Network Overview

As stated previously, on top of the physical media, the connectivity required by LTE can be realized by using different service types. All these traffic types are forwarded from the last-mile through the aggregation tiers to the core.

The most common implementation scenario including last-mile/aggregation and 3rd party network segments is displayed in Figure 6:

Figure 6. Transport network scenario based on Carrier Ethernet with L3VPNs.

The scenarios presented in Figure 6 have been defined only on top of Ethernet (IEEE 802.3), as it is the dominant layer 2 technology. The use of Ethernet interfaces has also been assumed for all base station and mobile core nodes. More detail on the required transport protocols can be found in Section 4.4.

In the 3rd party network segment, an IP/MPLS-based transport is used to implement a L3 VPN on top of Ethernet. The use of a L3 VPN allows the NaaS operators nodes to communicate privately over the 3rd party network, even though it routes through multiple 3rd party elements. Detail on this implementation is out of the scope of this module as this is a service provided by the 3rd party network.

2.3 Transport Technology Overview

The implementation of 4G Long-Term Evolution (LTE) imposes some requirements on transport networks such as more network capacity and latency reduction. These requirements are better served through terrestrial technologies (fiber optic and microwave). However, in rural areas, this becomes a challenge because satellite transport is usually the only feasible technology. For this reason, the transport network infrastructure is an essential component of the NaaS operator network.

An overview of commonly used technologies in the transport network is presented in the following subsections.

2.3.1 Fiber Optic Technology

Fiber Optic is the mainstay wired transport in mobile operator networks whenever it is available because of the significant inherent bandwidth carrying capability. Additionally, several additional techniques can be used to offset any bandwidth constraints and essentially render the fiber assets future-proof.

While fiber optic has tremendous operational capacity, its main limitation is the cost and logistics of deployment. Moreover, it can take several months to provision each cell site with fiber-optic transport.

The fundamental aspects of fiber optic technology design are discussed in the Primer on Fiber Optic Technologies Principles document.

2.3.2 Microwave Technology

Most operators heavily rely on microwave transport solutions in the 5 GHz to 80 GHz bands (in rural environments frequencies above 15GHz are not generally feasible). Microwave is a low-cost option for mobile transport, as it can be deployed in a matter of days and support a range of up to several tens of kilometers.

The main limitation of microwave links is the requirement of Line of Sight between transmitter and receiver. This represents an important constraint for its implementation, especially in rural areas due to steep conditions in the terrain. In addition, in many cases microwave requires a license to operate. Finally, in high frequencies, microwave links are subject to atmospheric effects or rain fade, which can attenuate the signal and limit its range.

The fundamental aspects of microwave technology design are discussed in the Primer on Microwave Technologies Principles.

2.3.3 Satellite Technology

Satellite technology is deployed in fringe areas of the network, usually in rural scenarios in emerging markets. Furthermore, satellite technology may be deployed as a temporary measure as the operator waits for regulatory microwave licenses to be approved. Satellite coverage is determined by satellite footprint and location-specific geographical and meteorological conditions.

The cost of satellite terminal equipment located at the base station is in line with microwave equipment. However, the OpEx burden generated by the satellite service fee can impact on the business case.

The fundamental aspects of satellite technology design are discussed in the Primer on Satellite Technologies Principles document.

3 Functional & Operational Requirements

Examination of functional and operational characteristics of the Transport & IP Network, providing guidance and general criteria to establish requirements for network designers and solution providers.

3.1 Transport Network Planes

As stated in Section 2.1.1, each eNB requires connectivity to the core elements and additionally for network management and potentially synchronization. The subsequent sections describe the different planes that must be supported by the transport network. More detail on the required transport protocols can be found in Section 4.4.

3.1.1 User plane

The user plane (U-plane) in the LTE transport network consists of the S1-U plane that is used to transport user data between the eNodeB and the S and P-GW using a general radio service tunneling protocol user (GTP-U). It also includes the X2-U plane used to provide connection to neighbor eNodeBs.

The protocol stack required to support the user plane is shown in Figure 7.

Figure 7. LTE user plane protocol stack

3.1.2 Control Plane

The control plane (C-plane) in the LTE transport network is composed of the S1-MME traffic that is used to transfer signaling information between the eNodeB and MME. This information is used for S1 bearer management, mobility and security handling and for application signaling messages between the user equipment (UE) and MME. It also includes the X2-C plane used to provide connection to neighbor eNodeBs.

The protocol stack required to support the control plane is shown in Figure 8.

Figure 8. LTE control plane protocol stack

3.1.3 Management Plane

The Management plane (M-plane) in the LTE transport network is the interface between the eNodeB and the O&M system. The M-plane consists of all data required to manage and monitor the status of the network elements. The volume of this traffic has to be considered during the capacity analysis and the frequency to collect these data.

The protocol stack required to support the management plane is shown in Figure 9.

Figure 9. LTE management plane protocol stack

3.1.4 Synchronization Plane

The Synchronization plane (S-plane) in the LTE transport network is important to ensure the proper functionality of the eNodeBs. The types of synchronization methods of interest for a mobile network are phase/time synchronization and frequency synchronization. More detail on the synchronization requirements is given in Section 4.7.

However, the implementation of this plane can be optional if it is decoupled from the transport network by using GNSS (global navigation satellite system) solutions at the eNB.

3.2 Availability & Resiliency

Most of the design considerations for the transport network architecture are constrained by the ability of the network to meet the proper levels of availability (the proportion of time a system is in functioning condition) which is supported by means of resiliency (the ability to readily recover). Table 1 shows the downtime per year according to the availability value.

Availability

Downtime per year

99%

3.65 days

99.9%

8.76 hr

99.99%

52 min

99.999%

5 min

Table 1. Availability and corresponding downtime per year.

The NaaS operators must define different availability levels for the transport network according to network segment (last-mile / aggregation). Figure 10 displays typical values of availability and resiliency requirements across the transport segments.

Figure 10. Availability on each segment of transport network

As it can be seen in Figure 10, the availability requirements of the transport network are derived from the availability requirements of the end-user service. A typical requirement for the transport network in a rural environment could be an availability value of 99.95%. In most cases the availability is dominated by the last-mile link.

Recommended values for the different segments within the rural NaaS transport network are presented in Table 2.

Service Level

Availability (%)

Last-mile Link

99,9%

Aggregation Link

99,99%

3rd Party Network

99,99%

Satellite Link

99,9%

Table 2. Availability Requirements Definition Example

3.3 Quality of Service Support

Quality of service (QoS) management is essential to distribute limited resources between different traffic classes. The concept of QoS used in LTE networks is based on classes, where each type of traffic (e.g. data, voice, signaling) is assigned a QoS Class Identifier (QCI) by the network to ensure adequate QoS for carrier traffic in the LTE network.

To allow separation of traffic in the transport network, translation of the radio QCI level into the transport QoS level is performed initially on the eNodeB. However, additional translations may be performed by intermediate transport devices along the path. For this reason, the QoS parameters in all the devices across the path must be correctly configured.

To allow the traffic separation, the IETF introduced the notion of Differentiated Services (DiffServ) classes as a Layer 3 mechanism. The key idea is to mark the priority traffic so that it is not affected by possible congestion due to the fluctuation of the link usage level. The concept of Differentiated Service (DiffServ) is a priority marking scheme carried in the IP header.

The Layer 2 QoS Mechanisms are used when the traffic traverses L2-only devices (e.g. a microwave radios). These devices use two 2-byte fields, Tag Protocol Identifier (TPID) and tag control information (Tag Control Information or TCI) to mark packets with the corresponding type of service.

Figure 11 displays the QoS Mechanisms used in LTE Network. The traffic type at the radio level is managed by the QoS Class Identifier (QCI). The operator defines the rules in the eNodeB by means of a QCI mapping table that contains a one-to-one mapping from the QCI value to the DSCP flag value and L2 QoS codes. Then, the intermediate device uses the QoS values to treat the traffic according to their capabilities (i.e. L2 or L3 QoS Mechanisms).

Figure 11. QoS Mechanisms in LTE Networks

It is recommended that NaaS operator ensure that traffic is classified and marked as close as possible to its origin, in most of the cases these tasks are performed by the eNode B for the Uplink traffic and by the Core elements for the Downlink traffic. Additionally, all the intermediate nodes must maintain the QoS markings. Finally, it is highly recommended that at least 4 different classes of traffic are defined to be used on the transport network.

4 Transport & IP Architecture Design Considerations

Analysis of multiple implementation options to be considered in the overall solution. In addition, methodologies for the most appropriate selection are provided.

4.1 Transport Technologies Selection

This section describes the architecture of different transport technologies to be used on the transport network. Moreover, it describes implementation alternatives for each technology and general guidelines to select the most appropriate.

4.1.1 Fiber Optic Transport Considerations

The general architecture using Fiber Optic as a transport technology is displayed in Figure 12. In this scenario, the enodeB is connected to a router located at the site, commonly named as Cell Site Gateway (CSG), which provides the transport services. The CSG is connected through a Fiber Link to the aggregation Node located in the transport node, which sends the traffic to the Mobile Core. Alternatively, an L2 switch can be deployed on the cell site location, moving the CSR location to the aggregation node.

Figure 12. Transport Architecture using Fiber Optic

For the implementation of the fiber optic architecture, three scenarios are identified for NaaS Operators:

  • Fiber Construction Implies building the fiber infrastructure up to the RAN site. Thus, the design of the fiber laying must be performed.
  • Fiber Link Leasing – Fiber links can be leased to a 3rd party with a presence in the region.
  • No Fiber Links Fiber optic may not be feasible if the cost of fiber infrastructure is out of the budget (e.g., high cost of fiber deployment in rural areas), and a 3rd party infrastructure cannot be used (e.g. no 3rd party with coverage in the area). In this case, the technology should not be considered in the later phases of network design.

In the following sections, the primary considerations for the two first scenarios are examined, and engineering guidelines are provided accordingly.

4.1.1.1 Fiber Construction

This scenario considers creating a fully dedicated, private physical network infrastructure by the NaaS operator. This approach can easily scale to handle whatever bandwidth requirements arise in the future.

On the other hand, the construction cost in rural areas is high, which may impact the overall CapEx project. Additionally, construction time can be lengthy and requires extensive planning and project management.

The principal considerations to support the fiber infrastructure design process are described in the following subsections.

Right of Way Definition

A right of way provides the right to pass across the lands of another, usually in a strip, acquired for or devoted to building facilities such as roads, railroads, or utilities. In this sense, rights of way are closely interrelated with both civil work for fiber infrastructure and technical engineering for network planning, operation and maintenance.

For the reasons stated above, a Right of Way Definition must be done to prioritize the types of road (e.g. main roads, highways, electrical lines) that may be considered during later phases in the transport network design. For example, a NaaS operator can only be granted to construct fiber infrastructure in rural roads, so other types of roads (e.g. highways) must not be considered during the design phase.

Fiber Optic Maximum Allowed Distance

It is the maximum distance allowed for a new deployment of fiber to be considered as feasible for implementation. This value must consider the technical conditions (e.g target availability) as well as commercial considerations (e.g maximum allowed CapEx for fiber deployment).

Fiber Optic Maximum Allowed Distance value is important for the overall transport network design as it directly affects the number of feasible fiber links. The greater the value, the greater the amount of fiber infrastructure to build, which increments the cost of deployment.

Fiber Optic Interfaces

In this scenario, multiple fiber strands are available as part of the fiber infrastructure laying, this characteristic allows to determine the fiber optic interface to be implemented. The fiber optic interfaces are modules used for the transmission of signals over the fiber optic medium. The complexity and cost of these modules varies according to the capacity supported (e.g. 1G, 10G, 40G or 100G).

Additionally, a Capacity Safety Margin must be defined to determine the capacity of exhaustion in the interface. For instance, a Safety Margin of 20% indicates that the interface capacity is exhausted when the real utilization hits 80% of the interface capacity. In this case, an upgrade in the interface must be considered.

Preferred Type of Fiber Laying

A new fiber deployment could require buried cables, aerial cables or underwater cables. Moreover, cable may be in conduit, innerduct or direct buried, aerial cables may be self-supporting or lashed to a messenger.

In rural areas, aerial cable is the most common fiber laying deployment due to its low cost and fast implementation.

Fiber Construction Summary

Table 3 provides to the NaaS Operator an initial set of recommended values for the Fiber Construction scenario.

Characteristic

Engineering Guideline

Right of Way Definition

– New deployments must prioritize the use of main roads (e.g. highways) and consider secondary roads as subject to permissions grant.

– The use of the electrical grid must be considered as priority if available.

Fiber Optic Maximum Allowed Distance

The maximum allowed distance between eNodeB and transport Node is generally 10km depending on the availability calculation and commercial agreement. For example, in a rural environment where a fiber cut occurs each 8 years per km and a MTTR of 1 day, the maximum allowed distance is 3km to maintain an availability of 99.95%.

Fiber Optic Interfaces

– 1G interface per Last-mile Links
– 10G interface per Aggregation Links
*Capacity Safety Margin of 20% of interface capacity

Preferred Type of Fiber Laying

Prioritize the use of aerial fiber.

Table 3. Engineering Guidelines for Fiber Construction Scenario

4.1.1.2 Fiber Link Leasing

NaaS operators can rely on an existing fiber deployment by leasing the fiber links to a 3rd party. This process implies an additional OpEx but avoids the investment in infrastructure deployment.

It is worth to say that the operation and maintenance activities of the fiber infrastructure is the responsibility of the 3rd party, lightening the operational workload. The monitoring to assure the SLA compliance can be done over the demarcation node between the NaaS operator and the 3rd party network. This approach is commonly known as Lit Fiber Leasing.

Another option for the NaaS operator is to only lease the fiber infrastructure of the 3rd party and install its own equipment. The approach described is commonly known as Dark Fiber. In this case, the NaaS operator can upgrade the capacity by changing the equipment at the ends of the fiber, becoming the link future-proof. Moreover, the fee cost remains fixed and independent of the amount of coursed traffic. Due to the nature of this scenario, some of the considerations of the previous section are applied.

In the following subsections, the principal considerations when leasing fiber links (Lit Fiber Leasing) to a 3rd party are described.

Fiber Link Capacity

In the Fiber Construction scenario, the capacity of fiber link is not a constraint as several fiber strands are available, and capacity exhaustion is not likely. However, the expected transmission rate on a leasing fiber link must be negotiated within the Service Level Agreements (SLAs).

The capacity of the fiber link must be appropriately forecasted to avoid surpassing the contracted limit. If the contracted capacity is exceeded, the traffic may be discarded, causing service degradation. Additionally, a Capacity Safety Margin must be considered in the link capacity.

Furthermore, in subsequent design phases, this value determines the maximum number of aggregated sites within a single transport link depending on the traffic generated by RAN equipment.

Fiber Link Availability

The availability in a leased fiber link is another parameter to be negotiated within the Service Level Agreements (SLAs). This value varies according to the network segment.

Fiber Link Leasing Summary

The typical values recommended for the NaaS operator regarding the Fiber Link Leasing scenario are summarized in Table 4.

Characteristic

Engineering Guideline

Fiber Link Capacity

– 100Mbps – 1Gbps per Last-mile Links
– 1Gbps – 10Gbps per Aggregation Links
*Capacity Safety Margin of 20% of link capacity

Fiber Link Availability

– 99,9% per Last-mile Links
– 99,99% per Aggregation Links

Table 4. Engineering Guidelines for Fiber Link Leasing Scenario

4.1.2 Microwave Transport Considerations

The general architecture using Microwave as a transport technology is displayed in Figure 13. In this scenario, the eNodeB is connected to the radio equipment located in the site, which is connected through a MW Link to the radio equipment located in the transport node. Then, a connection to the aggregation node is reached which sends the traffic to the Core elements.

Figure 13. Transport Architecture using Microwave

Microwave links may not be feasible due to unfavorable terrain conditions (e.g. mountainous terrain) or not suitable civil infrastructure (e.g. low rise towers). In this case, the technology should not be considered in the later phases of network design.

If the NaaS Operator selects microwave links as a suitable option, they should take the following considerations to support the microwave design process.

Frequency Band Selection

When selecting the operational frequency band to be used in the Microwave links, there are two main categories: licensed and unlicensed. These terms refer to the radio frequency spectrum characteristics set by the specific national government regulatory body. Licensed products require regulatory approval before deployment while unlicensed products can be deployed without any regulatory approval.

A comparison between the use of licensed and unlicensed frequency bands is presented in Table 5.

Characteristic

Unlicensed Bands

Licensed Bands

Licensing Cost

No licensing cost

License cost, in some markets spectrum charges can be high.

Total Time to Deploy

Very low as no license needs to be applied for and be granted.

High, as license needs to be applied and granted.

Interference Level

Medium to High due to external systems using the same frequency, causing link performance and throughput to be compromised.

Very low interference and no associated performance degradation.

Link Capacity

Low-Medium (~100-200 Mbps) due to limited available spectrum (typically 40MHz, but some equipment support up to 80MHz with some sensitivity constraints)

High (~1Gbps) due to larger communication channels (up to 112MHz) and reduced interference

Latency

High (~2-10 ms) due to time-division duplex (TDD) nature of most of unlicensed link (link cannot transmit and receive simultaneously)

Low (~0.1ms) due to use of frequency division duplex (FDD) links

Equipment Cost

Generally low due to mass-market WiFi chipsets

Higher, as specialized chips are required.

Table 5. Comparison of licensed and unlicensed frequency bands in MW

The final selection of the frequency bands to be considered in later design phases depends on the available budget and general capacity requirements. However, it is highly recommended to select a small number of frequency bands (e.g. two frequency bands, one for small distance links and one for large distance links) to reduce complexity in the design phases.

It is worth noting that licensed and unlicensed microwave technologies are not mutually exclusive. This means that NaaS operator can decide to use unlicensed bands for links with small distance and low capacity requirements whereas use licensed bands for long-range links and high capacity requirements.

Microwave Link Capacity

The Microwave Link Capacity is the expected transmission rate on a microwave link which must be defined for each of the network segments (last-mile and aggregation).

A link design must be performed in order to obtain the specific values (e.g. Modulation, Tx power) to reach the expected capacity. This process is detailed on the Tx LLD Module. Additionally, a Capacity Safety Margin must be defined to determine the capacity of exhaustion in a microwave link.

Allowed Topologies

Figure 14 displays different physical topologies that can be supported by using microwave links.

Figure 14. Microwave Topologies

The maximum number of sites that can be aggregated in a topology depends on the following characteristics:

  • Aggregated traffic on the link: The higher the capacity per eNB, the fewer eNBs can be aggregated. In addition, depending on the aggregated traffic, the last hop to the transport network may require the deployment of an aggregation link.
  • Target availability. The higher the availability target, the fewer number of hops that can be deployed in the microwave topology. One alternative to increase the availability is to implement the ring topology as it provides path redundancy; however, the number of microwave links may increase.

The methodologies to validate the feasibility of different physical topologies are included within the Tx HLD Module.

Microwave Transport Summary

The typical values recommended for the NaaS operator regarding the Microwave Link Deployment scenario are summarized in Table 6.

Characteristic

Engineering Guideline

Frequency Band Selection

– Unlicensed 5GHz for distances < 5km
– Licensed 15GHz for distances in the range of 5km-15km
– Licensed 7GHz for distances in the range of 10km-25km

Microwave Link Capacity

– 50Mbps – 100Mbps per Last-mile Links
– 100Mbps – 250Mbps per Aggregation Links
*Capacity Safety Margin of 20% of link capacity

Allowed Topology

– Star, chain and tree topologies.
*The feasibility of each topology must be validated during HLD Design.

Table 6. Engineering Guidelines for Microwave Link Deployment Scenario

4.1.3 Satellite Transport Architecture

The Architecture using Satellite as a transport technology is displayed on Figure 15. In this scenario the eNodeB is connected to a Very Small Aperture Terminal (VSAT) located in the site, which is connected through a Satellite Link to the terrestrial Satellite Hub. From this point, a connection to the aggregation node is reached which in turn, sends the traffic to the Core elements.

Figure 15. Transport Architecture using Satellite

The most common implementation scenario for satellite technology is to use a Satellite Service provider. The principal considerations when selecting and negotiating with a Satellite Service Provider are described in the following subsections.

Type of satellite orbit

Table 7 presents a comparison among different types of satellite orbits.

Orbit

GEO

MEO

LEO

Distance

35,800 km

2,000km-35,000km

160km-2,000km

Latency

250-500 ms

60-250 ms

30-50 ms

Throughput

Up to 500 Mbps

Up to 800 Mbps

1Gbps+ *(See NOTE)

Handover Periodicity

No HO

2-3 hours

15-20 min

Table 7. Satellite Orbit Comparison

*NOTE: As LEO systems have not been launched commercially, the 1Gbps value must be used as a theoretical reference only.

Despite the latency constraints, GEO orbits are preferable as they provide larger footprints and no satellite tracking mechanisms or handover are required. Additionally, LEO systems have not been launched commercially but may become an attractive solution when available.

Satellite Frequency Band

The satellite frequency band determines the available bandwidth and potential throughput in the satellite link. Ka-band (~30GHz uplink & ~20GHz downlink) and Ku-band (~14GHz uplink and ~12GHz downlink) are the most used bands for mobile transport.

Most Ka-band satellites offer 10x the throughput available on traditional Ku-band satellites. The increment of offered throughput also decreases the fee price of the satellite link. Moreover, Ka-band antennas have higher gain which results in less expensive VSAT equipment.

On the other hand, the higher frequency, the more a signal is susceptible to rain fade. That means that adverse weather conditions impact the Ka-band much more than at lower frequencies

For the reasons stated above, Ka-band is preferable to be implemented as the frequency band for satellite transport technology. However, Ku-band can be the only feasible option in extreme meteorological territories and in case of lack of Ka-band coverage.

Satellite Link Capacity

The Satellite Link Capacity is the expected transmission rate on a satellite link which must be defined. This value should be used as reference to negotiate the Service Level Agreements (SLAs) with the satellite service provided.

Additionally, a Capacity Safety Margin must be defined to determine the capacity of exhaustion in a satellite link. For instance, a Safety Margin of 20% indicates that the link capacity is exhausted when the real utilization hits 80% of the link capacity.

Satellite Transport Summary

The typical values defined for the Satellite Link Leasing scenario are summarized in Table 8.

Characteristic

Engineering Guideline

Type of satellite orbit

Preference on GEO Satellites

Satellite Band

Ka Band is preferred

Satellite Link Capacity

– 5Mbps – 15Mbps Downlink
– 1Mbps – 5Mbps Uplink

Table 8. Engineering Guidelines for Satellite Link Leasing Scenario

4.2 Transport Service Providers Evaluation and Selection

NaaS operator can select to rely on a 3rd party network when there is a well-established transport service that can provide the expected transport services with the corresponding Service Level Agreements (SLA). The complete process to include a transport service provider is included within the Procurement Module. This section focuses on the technical aspects to perform the evaluation of transport service offers.

There are two possible scenarios to include a 3rd party network:

  • For last-mile link implementation. This particular case only applies when the 3rd party provides the infrastructure from the RAN site up to the core element.
  • For aggregation link implementation. This is the most common scenario, where the NaaS operator deploys the last-mile link up to a 3rd party aggregation node. Then, it is the responsibility of the 3rd party to transport this traffic to the core elements.

From the technical perspective, Table 9 displays the typical requirements that a transport service provider must satisfy which must be defined according to required network service (e.g., last-mile or aggregation link implementation scenario).

Evaluated Characteristics

Availability

Reliability and uptime

Time-to-Repair

Time-to-Deploy

Capacity

Guarantee Capacity

Latency

Guarantee Latency

Security

Security mechanism

Table 9. Typical characteristics to be evaluated in a transport service provider.

The final decision to include an external transport service provider is also affected by the financial constraints of the project. By leasing an external network, considerable investment in infrastructure might be avoided (Capital Expenditures) and the operation and maintenance activities become responsibility of the 3rd party, lightening the operational workload. However, an additional OpEx must be considered along the lifespan of the network.

4.3 IP Networking Design

Figure 16 displays the specified 3GPP connections between eNodeB and core elements that are carried over the transport network and the specified protocol stack.

Figure 16. 3GPP Logical Interfaces Protocol Stack

In order to support the scenario presented in Figure 16, NaaS operator must define the different protocols and technologies to be used at different layers. The following subsections present the available options to be considered along the different layers considering IPv4 version protocol. Additional special IPv6 version protocol considerations are provided at the end of the section.

4.3.1 L1 / Physical Interfaces

Using Ethernet ports as physical interfaces reduces the need to install or upgrade physical ports as IEEE standards include data rates for 1Gbps (Gigabit Ethernet), 10Gbps (10G), 40Gbps (40G) and 100Gbps (100G). For instance, an initial 1Gbps port is adequate and may support necessary growth in the first years of the network.

In addition, Ethernet supports both electrical (twisted-pair cable) and optical fiber interfaces for physical connection.

Finally, the Ethernet physical port and the Ethernet L2 frame is basically agnostic to the higher layer protocol (i.e. protocol type field depends on the upper layer protocol). That means that for instance, IPv4 or IPv6 can be carried.

4.3.2 L2 / Data Link Layer

As the majority of LTE mobile equipment utilize Ethernet interfaces for transport, Ethernet based services are most suitable to be implemented in the transport network.

4.3.2.1 Virtual LAN (VLAN) Support

Virtual LANs are used to separate traffic logically or based on functionalities over the same Ethernet port (e.g logically separate O&M from the user/control traffic into a different VLAN). This is achieved by adding a tag (VLANs tag) to each Ethernet frame so that customer VLANs can be kept separate from provider VLANs. This tag is used by other Layer 2 devices to switch traffic in the network.

It must be noted that including a Virtual LAN (VLAN) tag increases the Ethernet frame header by 2 bytes, which needs to be taken into account when calculating Maximum Transfer Unit (MTU) at the Ethernet layer.Most base stations support a flexible way to bind eNB applications (S1/X2 U-plane, S1/X2 C-plane, Mplane, S-plane) arbitrarily to either: eNB interface address(es) or eNB virtual (loopback) address(es).

This flexibility allows base stations to be configured according to the transport services offered by the transport network, but also applying traffic separation (e.g. Management-plane from User/Control-plane traffic) as required. eNB interface IP address(es) can be assigned either to: one or more physical interface(s); or one or more logical interface(s).

A physical interface is typically provided by an Ethernet port, whereas a logical interface is provided by a VLAN termination. Different interfaces as well as VLANs belong to different IP subnets.

There are many configurations possible, but for the sake of simplicity only two exemplary configurations will be shown in Figure 17:

Figure 17. Example IP configurations of eNB

In the left example one IP address is used for terminating control, user and synchronization plane traffic and a second one solely for management plane traffic. As both IP addresses are running over different logical interfaces created by two VLANs, the addresses have to belong to different IP subnets. In the right example separate logical interfaces (VLAN) and IP addresses are used for each plane (user, control, management and synchronization).

The recommended configuration is to maintain separate logical interfaces and IP addresses for each plane. In this case, each plane will be mapped to a unique IP address in the network. A typical configuration of VLAN tags used to identify the traffic are presented in Table 10.

Plane

VLAN Tag

Control Plane (CP)

101

User Plane (UP)

102

Management (OAM)

103

Synchronization (SYN)

104

Table 10. Typical VLAN configuration

4.3.3 L3 / Routing Layer

LTE specifications define an IP based protocol stack for the logical interfaces (S1 and X2). Both versions of IP Protocol (IPv4 and IPv6) are included in 3GPP standards. Throughout the text IP mentions refers to IPv4, unless specifically mentioned.

Elements in the network need the next hop (router or interface) towards a given destination. This information is learned dynamically with the help of a routing protocol. However, 3GPP has no guidance of the use of routing protocols. An IP host (e.g. eNodeB) may have a static pre-configured entry for the default gateway only or it may run a routing protocol. The following subsections examine different approaches to implement the network routing mechanism.

Static Routing

The simplest method is the use of static routes. Their main advantage is that they are persistent regardless of any perturbation in the network, the state required is minimal and the absence of a control plane makes the device more scalable. However, the persistent state does not adapt to changes in the network topology.

The static route is recommended to be used on endpoints where a single connection is available. For instance, in a scenario where only one Cell Site Router is deployed, the eNode B can only send the traffic to this element. As soon as a second interface is added, some dynamic means of managing the state of the routing table on the device is required.

Routing Protocols

A routing protocol learns routes to destination networks from other routers in the network. This allows the elements to dynamically adapt to different network conditions (e.g. link failure) and is more scalable as network grows. One common classification of routing protocols is based on the role of the protocol in the network:

  • Interior Gateway Protocols (IGP) are used to exchange routing information within an Autonomous System (AS).
    The Open Shortest Path First (OSPF) protocol is a link-state routing protocol that uses a hierarchical routing model where a centralized routing domain is defined within an autonomous system. OSPF can be used in the aggregation segment when more complex topologies are presented.
  • Exterior Gateway Protocol (EGP) are used to exchange routing information between Autonomous Systems.
    The Border Gateway Protocol (BGP) is a distance vector that is used to advertise the route prefixes outside the organizational boundary without revealing the internal routing characteristics. BGP is commonly used to exchange prefixes assigned to mobile subscribers to the providers Internet peer.

Supporting protocols

Additional supplementary protocols included in the IP Stack must be considered during the implementation such as:

  • Internet Control Message Protocol (ICMP): ICMP is an integral part of the IP Protocol and all routers must support ICMP. With ICMP, hosts and routers can report errors and exchange diagnostic, control and status information.
  • Address Resolution Protocol (ARP): ARP is a L2 Ethernet broadcast which is asking for the MAC address of the next-hop IP address.

Figure 18, displays a typical implementation for L3 Routing protocols.

Figure 18. Typical L3 Routing Protocol Implementation

4.3.4 L4 / Transport Layer

The most commonly used transport layer protocols are examined in the following subsections.

User Datagram Protocol (UDP)

With the mobile transport application, UDP is the protocol on top of IP in user plane, and interfacing the radio layer protocols (GTP-U protocol on S1 and X2). UDP makes no assertion about the reliable delivery of data.

Stream Control Transmission Protocol (SCTP)

SCTP (Stream Control Transmission Protocol) provides reliable transport service for messages over the IP network. In the mobile transport network, it is used for signaling traffic. LTE signaling, S1 and X2 control plane, is over SCTP.

Transmission Control Protocol (TCP)

In the mobile transport protocols stacks, TCP is not included, since the user plane is over UDP/IP, and control plane over SCTP/IP. However, TCP may still be used in the management plane.

TCP is also used as a reliable transport mechanism for some of the IP and MPLS control plane protocols. For example, BGP uses TCP.

4.3.5 IP Protocol Stack Summary

The recommended protocols to be considered within the Reference Architecture are summarized in Table 11.

Description

Selected Protocol

L1 / Physical Interfaces

Ethernet Interfaces in all ports:
– 1G for last-mile interfaces.
– 10G for aggregation interfaces.

L2 / Data Link Layer

– Ethernet
– Support of 802.1q and 802.ad used in VLAN.

VLAN Definition

Four VLANs, one for each of the following planes:
– User plane
– Control Plane
– Management Plane
– Synchronization Plane

L3 / Routing Layer

– Support of IPv4 and IPv6 dual stack
– Static routing for a single site in the last-mile.
– OSPF in aggregation network
– BGP for connection to 3rd Party Networks
– Support for complementary protocols (ICMP, ARP)

L4 / Transport Layer

– UDP
– SCTP
– TCP

Table 11. Recommended IP Protocol Stack in Mobile Operator

Additionally, the Naas operator can use the Wizard for Tx & IP Architecture to support them in the definition of the required protocols to be implemented.

4.3.6 High Level IP Plan Design

In order to simplify the IP address management of the overall network, support several technologies and be able to perform an efficient troubleshooting, an IP Address Distribution Plan must be defined. An IP Address Distribution Plan is a document developed by the NaaS Operator that displays how the universe of available IP addresses will be distributed in a way that supports the required services. This plan should satisfy the following conditions:

  • It should be simple and follow logical rules.
  • It should be scalable up to the forecasted number of eNodeBs and network elements to be deployed in the network considering an additional margin for future expansions.
  • It should allow simple routing configurations and efficient route summarization.

The following subsections illustrate the methodology to generate the IP Distribution Plan using IPv4 version. A similar process can be performed when using IPv6 version considering that 16 octets are available. More detail on the specifics of this process can be found on the Primer on IP Planning Principles.

Network Range Definition

The first step to generate the overall IP Distribution Plan is to define the network range to be used to allocate IP addresses for each element in the NaaS operator network.

The Internet Assigned Numbers Authority (IANA) has reserved a number of IPv4 network ranges as private. These network ranges are reserved for organizations that want to build an internal network and are:

  • 10.0.0.0 10.255.255.255 (10.0.0.0/8)
  • 172.16.0.0 172.31.255.255 (172.16.0.0/12)
  • 192.168.0.0 192.168.255.255 (192.168.0.0/16)

The NaaS operator can select any of the above IP ranges to be used within their own network. However, it is highly recommended to use the 10.0.0.0/8, which allows the greatest flexibility.

IP Subnetting

Subnetting is the process used to partition of a network address space into more than smaller segments (e.g. divide one segment /8 into 16 segments /12). In this way, the NaaS operator can separate the network space addresses for each type of network elements (e.g. eNodeBs, transport devices, core elements).

In order to generate the IP Address Distribution Plan, there exist two primary conditions to consider: the type of network elements and the geographic distribution. In this sense, there are two main approaches:

  • Location-First: In this approach, contiguous subnets are allocated to each of operators defined regions, after which the subnet for a given region is divided among the different types of network elements. In this case, the most significant bits of the second octet are used to identify the geographical zone.
  • Element-First: In this approach, each type of network element is allocated with one subnet that is divided among the different regions. In this case, the most significant bits of the second octet are used to identify the type of network element.

Because an efficient route summarization for applications can be done using the Element-first approach, it is the recommended approach to generate the IP Distribution Plan supported by the NaaS operator.

Figure 19 displays a typical IP Distribution plan for a mobile operator. The displayed plan considers four different types of elements and it is assumed that the operator has divided its network in two geographic zones (North and South).

Figure 19. Typical IP Distribution Plan in Mobile Operator

Using the guideless presented in Figure 19, the IP Distribution in decimal notation is displayed in Table 12.

Type of element

Zone

Range

eNodeB

North

10.0.0.0/12

South

10.16.0.0/12

Router Loopback

North

10.32.0.0/12

South

10.48.0.0/12

MW_O&M

North

10.64.0.0/12

South

10.80.0.0/12

Core

North

10.96.0.0/12

South

10.112.0.0/12

Reserved

10.128.0.0/9

Table 12. Typical IP Distribution Plan in Mobile Operator

Additionally, Naas operator can use the High Level IP Distribution Plan Widget as support in the generation of the IP Distribution Plan.

4.3.7 IPv6 Considerations

The subsequent sections analyze the principal considerations regarding IPv6 protocol.

IP Addressing in Transport Network

One main driver in general for IPv6 is the unavailability of public IPv4 addresses. IPv4 addresses are a scarce resource, and several layers of NAT (network address translation) may be needed, which complicates the network and also introduces issues for many applications.

However, the IP addresses used for connectivity within the NaaS operator network do not need to be public (apart from some interfaces in the mobile core), and thus there is in general no pressing need for a larger address space. Thus, network elements within the transport network can use IPv4 addresses.

IPv6 Protocols in Transport Network

Even though both IPv4 and IPv6 are IP protocols, they are different protocols. IPv4-only devices are not capable of routing IPv6 packets. In order to support both protocols, network elements must support dual-stack configuration. The nodes that implement Dual-stack maintain two protocol stacks (IPv4 and IPv6) which work in parallel and, therefore, allow the end of the service or router to operate through any of the protocols.

Furthermore, separate versions of routing protocols must be maintained for each IP version (e.g. Open Shortest Path First version 3 (OSPFv3)).

Support for IPv6 Native Applications

In addition to the IP layer present in the mobile backhaul, end user applications use IP. The user IP layer is transparent to the transport network as it is encapsulated within the radio network layer protocols. User IP layer can be IPv6 even though the mobile backhaul is using IPv4. In other words, these two layers (end user IP and transport network layer IP) are isolated.

Figure 20 displays how IPv4 and IPv6 user applications can use the transport network independently of the implemented IP version. In this scenario, the devices that are required to support both versions of IP protocol (i.e. dual-stack devices) are User Equipment, eNodeB and Mobile Core elements.

Figure 20. Encapsulated user traffic over transport network.

In general, it is recommended to implement dual-stack network elements that can handle both versions of IP protocol (IPv4 and IPv6) in order to support legacy services and applications. This will allow the migration to a full IPv6 network.

4.4 Redundancy and Fault Detection Considerations

The present section includes the examination of diverse redundancy mechanisms, both physical and logical, as well as fault detection mechanisms to evaluate their applicability in the transport network.

Microwave Protection Scheme

The basic microwave protection scheme is the 1+1, which translates into radio redundancy, where two radio units create a single radio link. In the case of radio unit failure, the hot standby radio maintains the service on the site. The reason for implementing this type of mechanism is because recovery time may be lengthy in case of radio failure, especially in rural areas. Mean Time to Repair must consider the travel time to the site with the correct spare parts and the time it takes a qualified technician to climb and perform the replacement at height.

On the other hand, the main drawback of this mechanism is that two ratio units must be deployed for each microwave link which elevates the total CapEx of the deployment. For this reason, 1+1 protection scheme should only be considered in critical links (e.g. Microwave links in transport aggregation segment).

Ethernet Link Aggregation

Link aggregation supports combining multiple Ethernet links into a group, which is seen as a single link. The benefits are an increment in capacity as well as resiliency, as failure of a single link is tolerated. Furthermore, link aggregation also allows load sharing which is otherwise not supported by Ethernet.

Link aggregation group (LAG) can be implemented in aggregation segments as a redundancy mechanism to reach the target availability.

Router Redundancy Protocols

Router redundancy protocols (e.g Virtual Router Redundancy Protocol (VRRP) or Hot Standby Router Protocol (HSRP)) can be considered a form of L3 redundancy. The operational principle is to allow multiple routers to share a virtual IP and MAC address. There is one router elected as the master which assumes the role of forwarding the packets sent to the virtual IP address while the rest of the routers act as backup. In case the master fails, one of the backups becomes the master router.

Router Redundancy Protocols can be implemented in aggregation segments as a redundancy mechanism to reach the target availability. The specific selection of the protocol varies according to the selected Tx equipment.

Bidirectional Forwarding Detection (BFD)

Bidirectional Forwarding Detection (BFD) is a fault detection protocol that provides a low-overhead, short-duration method to determine a communication failure between devices and notify upper-layer applications (also called clients). BFD is a detection protocol that can be enabled at the interface and routing protocol levels.

For instance, a network running OSPF takes several tens of seconds to recover from a failure, using the default parameter settings. The main component of this delay is the time required to detect a failure using the timeout-based detection mechanism. In contrast, BDF typically detects a link failure in 50ms, which leads to a significantly decrease in network recovery time.

It is recommended to enable BFD in interfaces and routing protocol levels only on the elements located in the aggregation segment.

The recommended mechanisms to be considered within the Reference Architecture are summarized in Table 13.

Description

Selected mechanism

Redundancy mechanisms

Microwave Protection Scheme 1+1 in aggregation links

Ethernet Link Aggregation in aggregation links

Router Redundancy Protocols in aggregation nodes

Fault detection mechanisms

Bidirectional Forwarding Detection in aggregation nodes

Table 13. Engineering Guidelines for Redundancy and Fault Detection Mechanisms Selection

Additionally, the Naas operator can use the the Wizard for Tx & IP Architecture to support in the definition of the required mechanisms to be implemented.

4.5 QoS Mapping and Scheduling Considerations

In this section, a methodology to establish the general rules for the mapping between RAN QCIs and L3 QoS markings is presented. Additionally, different packet scheduling mechanisms are examined to determine their implementation suitability.

4.5.1 QoS Mapping Function

As stated in Section 3.3, the NaaS operator controls the Quality of Service by defining QCIs according to the characteristics of the service. In the transport network, a mapping among QCIs to DSCP bits and between DSCP bits and L2 p-bits are performed as illustrated in Figure 21. It should be noted that the ability to map QoS codes between layers is essential to ensure interoperability and providing end-to-end QoS.

Figure 21. Mapping Function of QoS values among different network layers.

Per-hop behavior (PHB) is an important concept to define the forwarding behavior applied to the different traffic types. A node performs the same PHB for packets with the same DSCP value. The NaaS operator implementation can use the following PHB types:

  • Expedited Forwarding (EF): EF traffic has the highest priority above all other traffic classes. The EF PHB applies to services that require a short delay, low jitter, and low packet loss rate, such as voice and signaling traffic.
  • Assured Forwarding (AF): Assured forwarding allows to provide assurance of delivery as long as the traffic does not exceed some subscribed rate. Traffic that exceeds the subscription rate faces a higher probability of being dropped if congestion occurs.
  • Best-effort (BE): The BE PHB focuses only on whether packets can reach the destination, regardless of the transmission performance.

The recommended mapping function among QoS codes of different layers is presented in Table 14. The table includes the per-hop behavior value which determines the treatment for each type of traffic.

Type of Traffic

QCI

p-Bit

DSCP

PHB

Data

9

0

0

BE

Voice

1

5

46

EF

eNB-Control

X

4

34

AF41

eNB Synchronism

X

5

46

EF

Management

X

3

26

AF31

Table 14. Recommended mapping function among QoS codes

4.5.2 Packet scheduling mechanisms

Packet scheduling mechanisms allow the congestion management within the transport network and must be implemented in any interface that can experiment congestion. Every time packets enter into devices faster than they can exit, there is a possibility of congestion. When congestion occurs, packets should be temporarily stored or queued in temporary storage for subsequent scheduling as shown in Figure 22.

Figure 22. Packet scheduling into different queues.

The commonly used types of queues are: Priority queuing (PQ) that are served in strict priority order and Weighted fair queuing (WFQ) that divides the interface bandwidth among the different flows according to their priority, ensuring a fair bandwidth distribution for all applications.

However, buffering memory is a limited resource on any interface. By queuing, buffers fill up, and packets can be discarded as they arrive, known as "Tail Drop"; or selectively dropped before all buffers are filled, known as weighted random early detection (WRED).

The recommended configuration for Congestion Management in the transport network for different types of traffic is presented in Table 15.

PHB

Assign Queue

Packet drop mechanism

EF

Priority queuing (PQ)

Tail drop

AF4

WFQ

WRED

AF3

WFQ

WRED

AF2

WFQ

WRED

AF1

WFQ

WRED

BE

Default Queue

RED

Table 15. Recommended configuration for Congestion Management.

Additionally, the Naas operator can use the QoS Mapping and Scheduling Mechanisms Template to support in the definition of the required mechanisms to be implemented.

4.6 Additional Architectural Design Considerations

This section provides guidance on the additional design considerations must be considered in order to support the overall Transport & IP Network solution.

4.6.1 Synchronization

Synchronization is fundamental in the LTE network operation as a failure in synchronization can result in spectral inefficiency and service degradation. The types of synchronization methods of interest for a mobile network are phase/time synchronization and frequency synchronization.

In FDD-LTE networks without special radio features may be deployed with frequency-only synchronization. In this case the synchronization signal is needed only for keeping the carrier frequency within required limits.

On the other hand, TDD-LTE and certain advanced LTE features (e.g. coordinated multi-point (CoMP)), require that all eNBs are phase synchronized in the sense that they must start radio frame transmission at exactly the same instant.

Global Navigation Satellite System (GNSS)

In satellite navigation systems (e.g. GPS and GLONASS), a GNSS receiver derives frequency and calculates time from the satellite signals, and the synchronization equipment then uses it as a reference for network timing.

The GNSS receiver can be a standalone device or embedded into the base station. A GNSS receiver can also be integrated into a collocated or nearby cell site router (CSR)

The implementation of a GNSS requires that the location of the GNSS receiver should allow an unobstructed view of the satellites of the used GNSS. Furthermore, the cost to deploy a GNSS receiver in each node may impact the overall CapEx deployment. For this reason, this synchronization mechanism can be considered when a satellite link is used as transport technology.

Precision Time Protocol (PTP)

The IEEE 1588 Precision Time Protocol (PTP) was developed in response to enable accurate distribution of time and frequency over packet-based networks. PTP requires the implementation of a server (PTP Grandmaster Clock) as the primary reference source for all of the PTP clients.

For the sites which do not have any GPS receiver, it is recommended to adopt PTP as clock synchronization protocol. The reason being it is standardized protocol, accurate and allows lower cost and the clock signal will be transmitted transparently through the transport network.

Synchronization Summary

The selection of synchronization mechanisms represents a trade-off among cost and complexity. For instance, for a small network deployment it may be easier to implement a GNSS receiver in all base stations and avoid the PTP infrastructure and configuration.

The default recommended synchronization mechanisms to be implemented in the transport network are presented in Table 16.

Description

Selected mechanism

Synchronization mechanisms

Global Navigation Satellite System (GNSS) preferred when satellite transport and unlicensed backhaul is used

Precision Time Protocol (PTP) for terrestrial transport

Table 16. Recommended synchronization mechanisms in the transport network

4.6.2 MTU Considerations

The maximum transmission unit (MTU) is the largest size packet that can be processed by nodes within the transport network. When a packet to be transported exceeds the MTU size of the egress interface, the packet needs to be fragmented in order to be sent forward. However, when MTU configuration does not match on the intermediate nodes, packets can be dropped generating a service degradation.

Because the connectivity between the eNodeB and the core elements uses GTP-U tunneling, the packets grow when transported over the transport network. Moreover, additional mechanisms required to provide the LTE service (e.g. VLAN tagging, IPSec) also increase the total packet size and fragmentation cannot be completely avoided. Thus, matching the path MTU with the IP packet size considering the additional headers (e.g. GTP, VLAN, IPSec headers) can reduce the possibility of fragmentation.

In Ethernet networks, the default value of MTU is 1500 bytes. In some cases, by using protocols that increase the frame header (e.g. VLAN, IPSec), mini-jumbo (3000 bytes) or jumbo frames (9000 bytes) may be required to prevent fragmentation.

It is recommended that NaaS operator determine the MTU to be used along the transport network considering all the additional headers resulting from architectural decisions. Furthermore, when the traffic traverses a 3rd party network, it is important to verify that the defined MTU value is set across every link in the external network to avoid traffic being dropped.

4.6.3 Security

This section presents the evaluation of different transport network security options to determine guidelines to select the most suitable implementation.

Network Element Security

One of the foundations for a secure network is to implement security at the network element level. This requirement applies to any node connected to the network.

Network element security covers multiple aspects of the node functionality, some of them need to be considered by the manufacturer at the early design and implementation phase, while others need to be considered by the NaaS operator during the deployment and maintenance phases.

The following list includes the most relevant actions that should be performed as part of the deployment and maintenance of the network element:

  • Support of user authentication and controlled access levels.
  • Support secure access by the management system user (e.g Secure Shell (SSH), Simple Network Management Protocol (SNMPv3))
  • Unnecessary physical interfaces are disabled.
  • Unnecessary services and ports are disabled.
  • Support of CPU Overload Control.
  • Default passwords are changed.
  • Passwords are regularly changed and only strong passwords are used.

Additional actions might be recommended by the transport equipment OEM.

IPSec Framework

IPSec is not a single protocol, but a collection of protocols and services that provide security for IP networks, including security protocols such as Authentication Header (AH) and Encapsulating Security Payload (ESP), Internet Key Exchange (IKE), and certain algorithms used for authentication and encryption.

IPSec provides secure services for IP packets mainly through encryption and authentication: User data encryption, Data integrity authentication, Data origin authentication and Anti-replay.

To achieve these services, the functionality of a Security Gateway (SEG) is required. This entity could be a dedicated network element or be included in other elements depending on the specific implementation. The typical deployment considers the SEG in front of the Evolved Packet Core (EPC). Figure 23 displays the architecture of the transport network implementing a dedicated SEG.

Figure 23. Transport architecture implementing IPSec framework

IPSec framework can be implemented in the following scenarios:

  • IPsec for protecting the LTE control traffic (S1-C, X2-C)
    This scenario should be considered by NaaS operators with self-built transport networks or with eNBs in areas secure enough to just encrypt the signaling traffic.
  • Full IPsec protection (control and user traffics)
    This scenario should be considered when the transport services are leased from a 3rd Party or when traffic traverses by untrusted networks.

The main drawback of this scenario is that the impact of IPsec is significant on the U-plane, as an additional overhead will be introduced. The additional overhead due to IPsec is about 14%. Furthermore, IPSec requires the implementation of more powerful hardware that is capable of encrypting and decrypting the traffic in real-time. This characteristic can increase the cost of network equipment.

Security of Management and Synchronization Plane

A secure mechanism to protect the Management plane (i.e. OAM traffic) must be available even if the transport network scenario is considered trusted by the NaaS operator. The Management plan can be protected by transport layer security /TLS mechanisms.

On the synchronization plane side, the following recommendations must be considered:

  • Handling the synchronization traffic in a separate and dedicated VLAN. If this option is not supported by the Tx equipment, this traffic can be transported together with the control flows.
  • In case synchronization is provided with PTP technologies, some 1588v2 inbuilt security mechanisms can be used.
  • It is not recommended to transport the PTP packets in an IPsec tunnel, as this has some impact on the delay variation

The recommended security mechanisms to be implemented in the transport network are presented in Table 17.

Description

Selected mechanism

Network Element Security

Mandatory Network Element Security Mechanisms in all nodes in the network.

User and Control Plane Security

Full IPsec protection when an untrusted 3rd party is used.

Management Plane Security

TLS protocol required for management plane

Synchronization Plane Security

PTP inbuilt security mechanisms with no encryption for synchronization plane messages

Table 17. Recommended security mechanisms in the transport network


5 Transport & IP Architecture Specification

Methodology to generate a final report that consolidates the outputs from this module:

  • Transport & IP Network Reference Architecture.
  • Transport & IP Network Requirements Specification.

The corresponding Templates are included as part of the Methods of Engagement.

5.1 Transport & IP Network Reference Architecture

The main deliverable is the Reference Architecture which contains the overall technical solution, engineering guidelines, technologies and concepts required to implement the Tx Transport network. The Reference Architecture shall contain the following aspects:

  • Overview: High-level description that summarizes the Tx Network Architecture.
  • Transport Technologies Selection: A description of the selected transport technologies to be considered as well as engineering guidelines to be used during Design phases.
  • IP Networking Design: A view on the protocols stack to be implemented in the different segments of the transport network.
  • Redundancy and Fault Detection Mechanisms: Description of the set of the Redundancy and Fault Detection mechanisms to be considered in the network implementation.
  • QoS Mapping and Scheduling Mechanisms: Description of the set of QoS Mapping and Scheduling Mechanisms mechanisms to be considered in the network implementation
  • Additional Architectural Considerations: A view on the additional architectural considerations (e.g. security and synchronization) to be considered in the network implementation.

5.2 Reference Architecture Generation

Generation of the Reference Architecture following the format and structure established. NaaS Operator can use the Transport & IP Reference Architecture Template as a reference to create their own version.


1. Network Monitoring Architecture Introduction

Once Network Architecture for the Radio Access Network (RAN), Transport and Mobile Core is defined, Network design and deployment takes place to finally enter the Operations phase. While operating the network, visibility into whats happening in the network is critical to identify problems, apply timely network and service repair/restore actions, improve performance, plan capacity and perform billing to customer Operators.

This module guides the NaaS Operator through an overview of network monitoring aspects required to define a solution. Furthermore, it guides the operator through the definition of these aspects, contributing to homologate the view across the organization and allowing for effective engagement with solution vendors.

The module addresses the high-level monitoring strategy, functional requirements, and implementation strategy, generating as final output the documentation of the Network Monitoring Reference Architecture and a formal Requirements Specification.

1.1 Module Objectives

This module will enable NaaS Operators to define all architectural aspects regarding Network Monitoring. The specific objectives of this module are to:

  1. Provide an information base and criteria to define the overall Network Monitoring strategy and data collection mechanisms to be implemented.
  2. Present an analysis of network monitoring functionality and a method to specify functional requirements for the Network Monitoring solution.
  3. Guide the NaaS Operator to establish an implementation strategy according to its own requirements and constraints.
  4. Provide guidelines to generate a consolidated Reference Architecture and Requirements Specification for evaluation and vendor selection for the Network Monitoring solution.

1.2 Module Framework

The Module Framework in Figure 1 describes the structure, interactions, and dependencies among different NaaS Operator areas.

The Strategic Plan & Scope provides business context for the NaaS Operator and serves as a guideline for the definition of the High-Level Network Architecture.

The Network Monitoring Architecture module is within the High-Level Network Architecture stream. The generated output of this module will serve as a reference for Network Design and will drive decisions in the Supply Chain Management.

Figure 1. Module Framework

Figure 2 presents a functional view of the Network Monitoring Architecture module where the main functional components are exhibited as well as its main inputs and outputs. Each functional block is discussed in its own section of the outline, detailing the impact of the respective inputs. In addition, a final section regarding outputs is provided.

Figure 2. Network Monitoring Functional View

The rest of this module is structured as follows: Section 2 provides an analysis and guidance to define the aspects related to monitoring strategy including data sources and monitoring technologies. Section 3 covers the functional requirements for a network monitoring solution; and Section 4 discusses the various options to implement network monitoring in a NaaS Network. Finally, Section 5 discusses how to tie all components together into a formal Reference Architecture and Requirements Specification.

2 Monitoring Strategy Definition

Overview of strategic aspects of network monitoring, including parameters, data sources and monitoring technologies, providing the NaaS Operator with criteria and guidelines to define its own Monitoring Strategy.

2.1 Network Monitoring Fundamentals

Network Monitoring is the practice of continuously measuring, analyzing and reporting outages, faults, deficiencies and performance metrics regarding network elements, links and services, contributing to maintain and optimize network availability and performance.

The process of network monitoring is automated through network monitoring systems, usually implemented as software only, but in certain cases using additional hardware to improve visibility into specific performance metrics.

Network Monitoring systems track and log network parameters and when predetermined thresholds are crossed, alarms are triggered and notified to the Operations personnel to initiate fault management procedures. In addition, performance metrics are also analyzed to look for optimization opportunities and to provide planning inputs.

Performance of Telecom business is greatly determined by network performance. In consequence, to improve/maintain business performance, network performance needs to be maintained/improved, requiring an effective network monitoring solution as you can only manage what you measure. In this sense, network monitoring is an imperative for any communication service provider, fixed or mobile and of any size.

An effective network monitoring solution provides the following benefits:

  • Improved uptime. Monitoring contributes to uptime in two ways: 1) detecting and fixing network problems before they cause outages, 2) accelerating the identification and solution of issues causing downtime.
  • Lower Operational expenses. Timely identification of performance degradations allows for planning of maintenance actions, optimizing costs. In addition, rapid identification of issues and solutions shortens the mean time to repair (MTTR), reducing the cost of field operations.
  • Lower churn. Better performance and improved uptime keep end users happy, reducing churn.
  • Higher Revenue. Better performance and improved uptime promote usage of the network. In addition, good performance makes the service attractive for new end users.

In the NaaS scenario, network monitoring is even more critical:

  • Smaller networks in comparison to country-wide Mobile Network Operators (MNOs). Each site represents a higher percentage of the overall network. Downtime may take a big chunk of revenue.
  • Higher time and cost to reach sites. Remote sites in rural environments may be hard to reach which makes it critical to minimize the number of unplanned site visits for corrective maintenance.
  • Limited Engineering, Operations and Field Resources. Efficiency for troubleshooting allows to keep resources focused on high-value tasks.

In the following sections, the NaaS Operator is guided through the definition of the main aspects of network monitoring.

2.2 Parameters & Thresholds

The first step towards the definition of the Network Monitoring Architecture is to identify what to monitor. To do so, the NaaS Operator needs to have a clear definition of the key performance indicators (KPIs) for its network. Then, the most important parameters to be monitored and thresholds for alerts are selected.

KPIs measure network performance on a macro level, they are metrics of interest for the NaaS Operator management team. KPIs are used to communicate the overall status of the network, benchmark networks, detect problem areas and manage Service Level Agreements with customer Operators and with Service Providers (e.g. transport service providers).

KPIs are relevant to network operations as they allow to:

  • Monitor and optimize the radio, transport and core network performance to provide better subscriber-perceived quality or better use of installed resources.
  • Rapidly detecting unacceptable performance in the network, enabling the NaaS operator to take immediate actions to preserve the quality of the network.
  • Providing the engineering team with the detailed information required for dimensioning and configuring the network for optimal use.
  • Troubleshoot and debug portions of the network

KPIs are calculated from network performance monitoring counters, alarms and status parameters, which are continuously collected by the Network Monitoring System. To achieve a well-monitored network, the following categories of KPIs shall be considered:

  • Availability
    The percentage of time that an element, link, network layer or service is available, i.e. it is in a state to perform its required function. Periods with high errors or with failures are considered as unavailable. Availability KPIs can be measured at the network element, link or service level.
  • Integrity & Performance
    Integrity refers to the degree to which a service is provided without excessive impairments. Performance KPIs measure how well a network or element can support the end user services, representing the quality experienced by the end user.
  • Load/Utilization
    Utilization is the percentage of resources being used from a network element or interface. This metric refers to any kind of resource: bandwidth, user capacity, CPU or memory. Utilization increases as more people demand resources from a network

In addition, for RAN and E2E the following categories can be considered:

  • Accessibility
    The ability of a service or resource to be provided, within specified tolerances and other given conditions, when requested by the end user.
  • Retainability
    The probability that a service or resource, once obtained, continues to be provided under the specified conditions for a specific period of time.

KPIs for each category and recommendations to set thresholds are provided in Table 1. In addition, the NaaS Operator is encouraged to use the KPI Specification Template to establish and document its own KPIs and thresholds.

Category

KPI

Description

Recommendations

Availability

Monthly Availability

Proportion of time that the service provided by the network is in working condition

E2E Availability per site should be at least 99% which represents almost 4 days of downtime per year. Availability of network elements and network sections should be adjusted accordingly

MTBF

Mean time between failures. Period with continuous uptime averaged over a month or a year

For information purposes only

MTTR

Mean time to repair. Period to repair a network failure averaged over a month or a year

To be defined based on SLAs with customers

Integrity & Performance

Latency

Round trip time. Average time for a packet to be sent, reach destination and get back to the sender

The E2E Latency or round trip time should be below 80 ms. Latency for each segment of the network should be adjusted accordingly. Latency for satellite links will be significantly higher (at least 400 ms)

Average Throughput

Average transmission rate for a specific service or interface. Should be measured in both ways: downlink and uplink

For RAN Equipment <50% of maximum eNodeB capacity.
For Transport equipment < 80% of transport capacity

Max Throughput

Maximum transmission rate for a specific service or interface within a time interval. Should be measured both ways: downlink and uplink

For information purposes only

Packet Loss Rate

Percentage of the number of lost or errored packets between two interfaces over a specified time interval

< 1%

Load / Utilization

Busy Hour Average Active Users

Average number of users with active sessions at busy hours

< 50% of eNodeB capacity

Busy Hour Peak Active Users

Peak number of active users at the busy hour

< 80% of eNodeB capacity

Traffic Volume

Total amount of data (in MB or GB) that has been transmitted at the user, cell and network level

For information and billing purposes. Can be calculated on a daily and monthly basis

Utilization

Percentage of resources being used at any point in time. Should consider cell, interface, CPU, and Memory utilization

80%

Accessibility

E-RAB Establishment Success Rate

Ratio of successful E-UTRAN Radio Access Bearer (E-RAB) establishment procedures with respect to the total number of E-RAB establishment attempts

99%

EPS Attach Success Rate

Ratio of successful Evolved Packet System (EPS) attachment procedures with respect to the total number of attach requests

5%

Call Setup Success Rate

Percentage of calls successfully established without facing blockage in the network as a ratio of the total number of call attempts

98.50%

Call Setup Time

Average time to establish a voice call, from call initiation to the called party being alerted

10 s

Retainability

E-RAB Drop Rate

Ratio of abnormal E-RAB releases with respect to the total number of E-RAB successfully established

2%

Call Drop Rate

Percentage of calls dropped due to technical problems or coverage gaps in the service providers network as a ratio of the total number of calls successfully established

1.50%

Call Success Rate

Percentage of calls successfully established without facing blockage in the network as a ratio of the total number of call attempts made and then successfully terminated from the user-end

97%

Table 1. Baseline KPIs and Recommendations.

Once KPIs are established, the NaaS Operator must define which parameters should be monitored throughout the network to feed the calculation of the KPIs and to identify any performance issues or element failures.

Table 2 shows a baseline recommendation of parameters to be monitored in the NaaS network with comments for applicability and further customization.

Network Element

Parameter

Comments

Generic

Interface Status

Physical interfaces at minimum. Logical interfaces may be monitored as well

Traffic Counters

Various granularities can be defined. Baseline is at the physical / logical interface level

Interface Utilization

For all interfaces

Packet Errors

Error counters for each interface/link

Interface Rx Power

Applies for Radio or Optical Interfaces

CPU Utilization

For all network elements

Memory Utilization

For all network elements

Network Element (NE) Temperature

For all network elements

Power Status

Whether the Primary, Secondary, and/or Batteries are working properly

Fan Status

Only for NEs with integrated fans

RAN Equipment

Throughput

Average and Maximum, Downlink and Uplink

Radiofrequency Resource Utilization

Number of PRBs being used or percentage of PRBs being used in Downlink and Uplink

Attached Users

Number of registered users

Active Users

Number of users with an active session

Radio Resource Control (RRC) Protocol Performance Counters

Includes number of E-RAB requests, attach requests, call setup requests, successful and failure events

Downlink Latency

Processing latency at the eNodeB, from the time a packet is received at the Backhaul interface to the time the packet is scheduled in the air interface

Transport Equipment

Layer 2/ Layer 3 Protocol Status

Refers to routing tables, neighbor status, convergence status, etc.

Layer 2/ Layer 3 Protocol statistics

Refers to statistics regarding number of events, number of neighbors, etc.

Internet Control Message Protocol (ICMP) Statistics

Ping statistics: number, RTT, packet loss

Core

Control Plane Performance Counters

Counters for signaling events

User Plane Performance Counters

Traffic statistics at the user plane that may be useful for billing customer MNOs

Power

Mains Status

AC Power received from the power company

Battery Status

Charging/Discharging, Voltage

Input Voltage

Input voltage to the power supply unit and rectifiers

Output DC Voltage

Output voltage delivered to network elements

DC Load Current

Current drained from the power supply unit to the network elements.

Table 2. Baseline parameters to be monitored in a NaaS Network.

In addition, the following recommendations are provided to the NaaS Operator:

  • Dont over-monitor. For the baseline, define critical parameters required by the KPIs and those that allow to quickly identify failures and performance degradations. Establish basic parameters for continuous monitoring and always verify that the monitoring system allows for additional parameters on-demand.
  • Remember that alarms can come from thresholds defined per parameter and from alarms generated by the network elements. In some cases, the alarms from network elements can be enough.
  • To set alert levels is important to identify meaningful thresholds that prevent constant alerts but allow for timely preventive or corrective actions. These levels need to be set as close as possible to the actual threshold of service-impacting trends or events.

In order to customize monitoring parameters and thresholds for its own network, the NaaS Operator may use the Monitoring Parameters Definition Wizard.

2.3 Data Collection Strategy

In the previous section, recommendations to specify network parameters to be monitored have been provided. Now it is vital to consider that network monitoring data is continuously generated and important decisions regarding where and how often to collect this data must be made.

As the amount of monitored parameters and collection frequency increases, operational improvements follow an S-curve. As shown in Figure 3, a small amount of data with low frequency offers low value, then, there is an inflection point where incremental data can deliver high value; however, there is a second inflection point from which more data starts to have diminishing returns.

Figure 3. Operational Improvements and Data Collection Strategy.

Furthermore, it is important to consider that a higher amount of data and higher collection frequency means higher monitoring complexity and cost, and higher monitoring overhead on the network. Thus, the NaaS Operator should find an equilibrium point where the operational improvements are exploited under an adequate investment and complexity for the monitoring system.

The following recommendations are provided for the NaaS Operator to define its own monitoring intervals throughout the network:

  • Availability for all the network elements and service components shall be configured with the lowest monitoring interval possible, either every minute or every 5 minutes. Depending on specific conditions and SLAs with customer MNOs, higher intervals can be specified.
  • The monitoring interval for traffic metrics will depend on the location and level of aggregation of each link. For last-mile links, 15 minutes should be enough. However, for aggregation links and interfaces to 3rd party networks, lower intervals, such as 5 minutes shall be used
  • CPU and memory utilization can be configured with a 15-minute interval for RAN and Core equipment, while 60 minute-intervals are acceptable for transport equipment
  • Other parameters shall not be continuously monitored, but triggered based on thresholds, such as L3 protocol status and power equipment metrics.

2.4 Network Data Sources

Typically, monitoring data is obtained from network elements. However, in some cases an additional level of visibility or functionality requires additional hardware to monitor the network. Furthermore, virtualized functions and cloud services cannot be directly monitored; instead, APIs are used to monitor their performance and availability status.

Below, a description and comparison of the most common network data sources are presented. Afterwards, guidelines to select the most suitable data sources based on the NaaS Operator use cases, constraints and requirements are presented.

  • Network Element Management Information Base
    The primary source of network data are network elements, which provide health metrics and events as well as traffic statistics or counters, according to a management information base or data model.

    A management information base (MIB) is a database that contains hierarchical information in a tree structure. The MIB of each device is accessed through the SNMP protocol (see Section 2.5), which also defines the basic format of the MIB. (All other MIBs are extensions of this basic management information base.)

    In addition to the SNMP approach that polls every monitored network element, there are other protocols that work with a push model where network elements send telemetry data to the monitoring system periodically or driven by events.

  • Network Element Flow Data
    Certain network elements (such as switches, routers, firewalls, etc.) have the ability to generate summarized data and characteristics of the IP conversations that go through them. These characteristics and performance data are embedded in fixed-size data records commonly known as flows that are periodically sent to a collector, which converts flow data into useful charts and graphs.

    This type of data is useful in cases where the NaaS Operator wants to obtain visibility into individual services and devices performance.

  • Network Element Mirror Port
    Port mirroring is a technique through which packets that enter or exit a physical or logical interface are copied and sent to a local interface for local monitoring or to a VLAN for remote monitoring.

    This technique is used when visibility at the packet level is required, usually for applications that inspect packets such as enforcing policies, detecting intrusions, monitoring and predicting traffic patterns and correlating events.

  • Terminal access point (TAP)
    Similar to the mirror port, the TAP monitors raw network packets. However, a tap is an independent device that is directly connected to the cabling infrastructure to make non-intrusive copies of the packets that are sent and received through a network link.

    The additional hardware may be justified when packet visibility is required and the overload of port mirroring is not acceptable for the monitored devices. In other words, a tap provides full visibility into what happens in a link with no impact to network elements at the price of additional hardware.
  • Probe/Synthetic Test
    Probes are devices that allow to perform active tests in the network from the location that they are installed. This device provides visibility and detail needed to activate new services, troubleshoot performance problems and identify their root causes.

    Another use case is SLA verification and QoE monitoring as the probes enable operators to emulate end user behavior, measuring the real performance of the network under the specific service conditions.

    These tests (synthetic tests) are often supported in specific network elements that can trigger and respond to industry-standard tests such as 802.1ag, Y.1731, Y.1564 and TWAMP.

  • Application Programming Interface (API)
    An API is a set of specifications and protocols to integrate two applications; it allows a product or service to communicate with others independently of how each service is implemented. APIs in the network monitoring context, are useful to obtain data from 3rd party sources.

    In particular, APIs are useful to monitor and communicate with cloud services and virtualized elements. In addition, data collection from intermediate systems such as Operation Support Systems (OSS) or Network Management Systems (NMS) that collect and process data directly from network elements should be done through APIs.

Table 3 summarizes and compares the network data sources described above.

Data Source

Type of Data

NE Impact

NE Support

NE – MIB/ Data Model

Availability Status & Performance Counters

Low

Supported by virtually ALL network elements

NE – Flow Data

IP Flow Statistics

Medium

Supported by carrier grade routing and switching equipment

NE – Mirror Port

Packet

High

Mainly routing and switching equipment. A physical port may be required on the monitored network element

TAP

Packet

None

Not Required

Probe & Synthetic Test

Performance Measurements

None

Certain network elements, may support synthetic tests

API

Availability, Performance & Status

None

Not Required

Table 3. Network Data Sources.

In order to select which data sources to implement for its Network Monitoring Architecture, the NaaS Operator can go through the process flow presented in Figure 4. Also, the Network Data Sources Selection Wizard can be used for the same purpose.

Figure 4. Network Data Sources Selection Process.

2.5 Monitoring Protocols & Technologies

In addition to data sources, various protocols and techniques exist to collect data from these sources. In the following sections, the main protocols and techniques are described, providing guidance for the NaaS Operator to determine which methods to implement based on the use cases and requirements for each of them.

2.5.1 SNMP (Polling)

Each network element has embedded agents that speak Simple Network Management Protocol (SNMP). These agents are interrogated with a polling-based approach, returning metrics from the network element.

Polling consumes many resources from the network monitoring system, especially as the number of monitored network elements increases. A problem with SNMP is that low latency updates coming from network elements may be delayed due to the polling mechanism

Use Case: SNMP covers the basic requirements for availability and status metrics, including slow traffic monitoring.

Requirements: SNMP is supported by virtually every possible network equipment and monitoring software. The only requirement to set it up is to perform an adequate configuration at the network element and monitoring system level.

A deeper exploration of SNMP features is given in the SNMP Primer.

2.5.2 NetFlow

NetFlow is a method used to report on the traffic that is passing through network elements. Rather than deploying probes around the environment to capture, analyze, and report on traffic behaviors, flows are collected and processed by collectors. The utilization of collectors allows to apply analysis technologies to provide insight into which devices and applications are consuming bandwidth, how long the conversations are lasting and who is participating in them.

Because the data is summarized (and sometimes sampled), a degree of detail is removed to reduce bandwidth overhead, simplify processing and to extract meaning from the actual network data.

Several protocols, both proprietary and standard-based that operate under this principle have been defined. Proprietary protocols include among others NetFlow from Cisco, J-Flow from Juniper and NetStream from Huawei. On the other side, standard-based protocols include sFlow which is considered an Industry Standard and IPFIX that has been specified by the IETF.

Use Case: Detailed traffic analysis where it is not required to analyze a specific set of network packets, nor does it require timing or delay information. Basically, NetFlow enables network planning and optimization with the right granularity to keep the system simple and with reasonable storage requirements.

Requirements: To leverage NetFlow technologies for network monitoring, the NaaS Operator needs to make sure whether network elements support it and under which protocol. In addition, hardware or software-based collectors are required as well as some form of pre-processing and long-term storage.

For a detailed description and comparison of NetFlow and Packet technologies, the NaaS Operator may consult the Primer on NetFlow and Packet Data.

2.5.3 Packet Collection

Packet collection consists of retrieving an exact copy or relevant data from every packet transmitted over a network interface. Packets offer full visibility into bursty traffic behavior, enabling a greater insight and precision than other technologies. Packets provide information about the precise microsecond when a packet passes a capture point, allowing to measure response time and network latency.

Packet collection is enabled through the potentially costly (and often, impossibly costly) implementation of hardware collectors to capture packets.

Use Case: Detect and troubleshoot performance problems that are beyond network congestion and that are timing-based such as excessive delays, latencies, etc. Usually applied only to critical interfaces such as mobile core interfaces to validate protocol signaling and procedures.

Requirements: A capture point (TAP) and capture device (collector) must be deployed to get packet visibility on specific interfaces to be monitored. As the network grows, more TAPs and collectors are required. When signaling traffic is encrypted, an additional burden is put into the collectors that must perform decryption of this traffic, implying additional processing capacity and costs. High storage requirements are also characteristic as information about each packet is stored for analysis.

The NaaS Operator is encouraged to read the Primer on NetFlow and Packet Data for further comparison of these technologies.

2.5.4 Streaming Telemetry

Similar to SNMP, with Streaming Telemetry, availability and performance data is collected from the network elements, but instead of using the polling mechanism, it uses a pushing mechanism to overcome the inefficiencies of polling. Under this pushing mechanism, the monitoring system subscribes itself to certain updates that each network element issues (streams) periodically or on an event basis.

By using the push mechanism, Streaming Telemetry enables a high reporting frequency (up to 10 s) for critical data and offers complete access to operational and configuration data using open data models. In addition, all parameters reported are time-stamped, allowing for a better understanding and analysis of performance metrics.

Streaming Telemetry can use various protocols for telemetry transport such as UDP, TCP and gRPC, which is a web-based protocol that significantly reduces the required bandwidth to transport data and adds TLS encryption of telemetry data.

Use Case: Streaming Telemetry is suited to any SNMP use cases and, in some cases, may replace NetFlow. However, the core use case is when low latency updates are required. In addition, it is used to retrieve configuration data and to interface with APIs that offer telemetry subscription services for cloud and virtualized network elements.

Requirements: The basic requirement is that the network elements or services must support the push mechanism and a suitable transport protocol. In addition, network elements should provide their data models in a standard format, being the YANG model the prevalent choice in the Industry. The Monitoring system architecture remains almost the same as with SNMP.

Further exploration of the principles, features, and benefits of this technology is provided in the Streaming Telemetry Primer.

2.5.5 Selecting Monitoring Protocols & Technologies

Based on the above descriptions, the NaaS Operator must define which monitoring protocols and technologies will be part of its Network Monitoring Architecture. The following recommendations are provided as a supporting guide for the NaaS Operator analysis:

  • SNMP is the basic monitoring protocol and must be considered as the baseline, at least in the early stages of the network operation.
  • Streaming telemetry can be considered when supported by network elements and for specific cases where advanced analytics or high update frequency is required. Another case where streaming telemetry might be required is when monitoring cloud services and virtualized network elements through APIs that offer subscription services.
  • NetFlow and Packet data should be considered under very specific circumstances. NetFlow, in cases where a detailed traffic analysis is required but no precise timing information is of interest. Packet data should be considered only for key interfaces and when detailed troubleshooting or call flows analysis is required.

As a complement to these guidelines, the NaaS Operator can make use of the Monitoring Technology & Protocol Selection Wizard.

3 Functional Requirements Specification

This section consists of a walk-through to explain functional aspects of network monitoring solutions and guidance to specify baseline requirements before the NaaS Operator engages in discussions with potential solution providers.

Furthermore, a Functional Requirements Template is provided to the NaaS Operator for customization, including a set of generic requirements that can be further complemented and edited according to each NaaS Operator network architecture and strategy.

3.1 Events & Alarms Processing

An event is something that happens in the network and has some importance or impact on the overall availability, performance or status. Most events are asynchronous due to changes in the state of network elements, links or user requests.

An alarm is a type of event, which is either generated automatically by network elements or by the network monitoring system based on performance thresholds.

In effect, one of the fundamental requirements for a Network Monitoring Solution is the ability to monitor events and alarms, and generate alerts for network elements, links and services. Events and alarms are monitored through syslog or SNMP traps, which are alert messages sent from a network element towards the network monitoring system. Furthermore, having continuous performance monitoring enables the generation of alerts based on performance levels, which is a vital tool to detect potential issues before they reach critical levels that ultimately disrupt network services.

As an alternative to SNMP or syslog, when using streaming telemetry, events and alarms based on performance can be configured at the network element to notify the monitoring system when they occur.

The NaaS Operator shall define events and alarms processing requirements for the Network Monitoring system. Minimum requirements are listed below and included in the Functional Requirements Template for customization:

  1. The System shall perform continuous real-time monitoring of network alarms using monitoring protocols and technologies defined by the NaaS Operator
  2. Vendor and network technology agnostic. The System shall support the monitoring of network elements (and cloud services) for any network technology (RAN, Transport, Cloud) and any vendor that supports the defined collection mechanisms.
  3. Alarm reporting. The System shall process and report, at minimum, the following types of alarms:
    1. Interface and Port failures
    2. Power source loss / power source switch
    3. Synchrony alarms
    4. CPU/Memory Overload
    5. Service outages
    6. Temperature
    7. External Alarms connected to network elements
  4. Threshold configuration. The System shall allow configuration of performance thresholds for alarm generation
  5. Alert prioritization. The System shall allow configuration of alarm criticality and prioritization of alerts.
  6. Alarm correlation and deduplication. The System shall identify interdependencies of alarms generated from an initial alarm, consolidating alarms that are duplicated.
  7. Alarm filtering. The System shall allow the operator to filter alarms for visualization according to several criteria. At a minimum, filtering by network element, geography and criticality.
  8. Alarm Disabling. The System shall consider an option to manually clear or disable alarms to minimize alerts during maintenance works or changes in the network.
  9. Automated Alarm Clearing. The System shall clear all alarms once the network condition returns to the normal state.
  10. Alarm notification (alerting). The System shall generate notifications to specific personnel or other systems (such as Ticketing Systems) through various channels such as SMS, email, messaging apps and API notifications.
  11. Alarm Log. The System shall keep a log of all the alarms presented in the network.

3.2 Analytics

Network analytics is the practice of using different types of network data to identify trends and patterns, and using statistical techniques to improve the performance, reliability, or security of the network as well as to find out solutions for critical problems.

Network analytics provide network operators in general and NaaS Operators in particular, with deep insights of their networks, enabling them to make better decisions for troubleshooting, threat remediation decisions, business planning and performance optimization.

Other benefits of an efficient implementation of network analytics in the network monitoring system are network scalability, reduced operational costs and resource optimization:

  • An operation can scale to many devices, clients, users, and applications, while improving overall user experience, without a substantial increase in operational costs as Analytics enables not only addressing enterprise network problems in real-time but also predicting where network problems may be brewing.
  • Analytics contribute to understanding the network, and balance and utilize resources in the best way possible so performance can be optimized with lower cost structures.

A fundamental aspect of analytics is to baseline network behavior over a couple of weeks or months. Knowledge of baseline behavior aids proactive troubleshooting and provides essential input for advanced analytics algorithms.

Network Analytics can be classified in various ways. For simplicity, only two classifications are presented here. The first classification divides analytics into Real-Time and Batch or Historical Analytics. Real-time analytics uses real time data to enable quick detection of network and service degradation as well as continuous network status reporting. On the other hand, Batch Analytics allows for powerful analysis for diagnostics and optimization that is done offline using historical data.

The second classification is based on the function performed by the analytics. As shown in Figure 5, there are Descriptive, Diagnostic, Predictive and Prescriptive Analytics and each of them answers a different question.

Figure 5. Network Analytics Categories.

  • Descriptive Analytics uses historical and real-time data to identify the current status and behavior of the network. This type of analytics can be used for SLA monitoring, Network Health Check, Performance Reporting, Fault Management and quality of experience estimation.
  • Diagnostic Analytics allows to find the drivers and causes of faults and performance degradation, using historical and real-time data. Applications of this type of analytics are root cause analysis, security analysis and non-service impacting network anomaly identification.
  • Predictive Analytics focuses on finding trends that ultimately allow to forecast network behavior, generating data on the future network status. Predictive Analytics can be used for fault prediction, schedule preventive maintenance, and traffic forecasting and capacity planning.
  • Finally, Prescriptive Analytics takes insights from current and forecasted network data to suggest actions in order to improve network performance and availability, such as load balancing, congestion control and network optimization. To fully exploit the potential of prescriptive analytics, the network must incorporate automation mechanisms.

In todays market offering, there are several monitoring solutions that incorporate Artificial Intelligence and Machine Learning into their analytics products. Thus, it is important to have a good understanding of what this means. Machine learning is a subset of artificial intelligence, and it gives analytics engines the ability to automatically learn and improve from experience without being explicitly programmed.

Machine learning offers significant increases in the accuracy of insights and remediation because the decision trees are modified to meet the specific conditions of a network’s configuration, its installed hardware and software, and its services and applications.

However, the NaaS Operator should be aware that Analytics are as good as the data that is fed to these tools. Thus, analytics should be appropriate to the network monitoring strategy. It is not recommended to have full Machine Learning Prescriptive Analytics if the monitoring of the network is going to be using SNMP and 15-minute reporting intervals.

In many NaaS scenarios, simple analytics have the potential to provide huge value for NaaS Operators at a reasonable cost. In consequence, the NaaS Operator shall define adequate analytics requirements for its Network Monitoring system. Minimum requirements are listed below and included in the Functional Requirements Template for customization:

  1. Baselining. The System shall have the capability to establish a baseline behavior of the network, notifying the Operator deviations from it and allowing to establish alarm thresholds.
  2. Real-Time & Batch Analytics. The System shall provide real-time and batch analytics, each with different processing software or mechanisms, supporting real-time alerts and performance reporting, as well as historical performance analysis
  3. Descriptive Analytics. The System shall support descriptive analytics that proactively detects network and subscriber performance degradation, anomalies and congestion and allow for calculation of the KPIs specified by the NaaS Operator.
  4. Trending Analysis. The System shall provide trending features, allowing to identify potential performance degradations or security threats, detect potentially damaging conditions, including those that only occur during peak traffic conditions.
  5. Diagnostics. The System shall provide the diagnostic workflows and forensic data to identify the root causes of performance degradations or failures

Additional requirements that may be considered by the NaaS Operator, based on the size and maturity of the network, monitoring strategy and financial constraints are:

  1. Predictive Analytics. The System may incorporate machine learning algorithms for performance forecasting and fault prevention.
  2. Prescriptive Analytics. The System may provide prescriptive capabilities that allow the network to evolve from reactive detection to preventive automatic actuation in advance of network problems.
  3. Advanced Performance Analysis. The System may support detailed analysis of service and network performance to enable Quality of Experience estimation and to analyze LTE radio and signaling events, identifying potential issues with specific cells, users, applications and devices.

3.3 Dashboards & Reports

As stated before, you only manage, what you measure. However, once measuring is taking place, the next critical issue is to visualize. Data becomes useful only when it is presented clearly to the right audience that can be Operations personnel, Engineering personnel or Managers.

A network dashboard is a panel that provides an at-a-glance overview of the current status of the network, with KPIs and critical parameters from cell sites, microwave radios, satellite links, routers, switches, core entities, services and underlying infrastructure. A dashboard allows to see the big picture of the network and provides the ability to drill-down for specifics on quasi-real-time. Figure 6 shows an example of a network monitoring dashboard.

Figure 6. Network Monitoring Dashboard

On the other side, reporting is key to keep stakeholders aware and informed about the status and behavior of the network over a period of time, using historical data. Apart from the internal evaluation of the network, reporting might be required for SLA and regulatory compliance, and for customer billing.

Reporting is a task that is performed periodically or on-demand in every network. Thus, it should be automated to ensure that resources devote time to other high value-activities.

The NaaS Operator shall define Dashboarding & Reporting requirements for the Network Monitoring system. Minimum requirements are listed below and included in the Functional Requirements Template for customization:

  1. Network Visualization. The System shall provide configurable views of the network, considering at least a Geographic and topology view.
  2. KPI Visualization. The System shall be able to show the selected KPIs on a list, chart or map.
  3. Top Offenders. The System shall identify and show elements considered as top offenders under specific KPIs defined by the NaaS Operator.
  4. Deep Dive Analysis. From the dashboard view, the System shall allow deep-diving into a region, network technology, site, network element or interface.
  5. Alarms & Notifications. The System shall generate visible notifications of alarms through text and changing the color of icons in the dashboard depending on the criticality.
  6. Dashboard customization. Dashboards shall be configurable for various levels and roles, allowing to show different network views, KPIs, charts, etc.
  7. Report Generation. The System shall provide reporting capabilities, allowing to configure metrics to be reported, including KPIs, individual element parameters applying averages, distributions historical series and other operations.
  8. Reporting Granularity: Report granularity shall be customizable, including reports at the network, region, network segment, technology and network element level.
  9. Report Periodicity. The system shall support on-demand or periodic generation of reports, with configurable periodicity including at least daily, weekly, monthly and yearly options.
  10. Data Export. The System shall allow to export any monitoring data as files in machine-readable formats or through a web API.

3.4 Security

Security is a vital aspect in every network and system as attacks can disrupt the network and end user services, potentially impacting the privacy of the Organization, Customers and End Users.

In the case of the Network Monitoring system, security has to be addressed in two ways: Security in the connection to Network Elements and network data collection; and security and integrity of the Monitoring System itself.

In the following, the minimum set of security requirements that the NaaS Operator should specify for the Network Monitoring System is presented. These requirements are also included in the Functional Requirements Template, which can be customized by the NaaS Operator.

  1. Access Control Lists. Access to the network elements for monitoring shall be controlled through access control lists.
  2. Secure Protocols. Data collection from network elements shall be performed using secure protocols. If using SNMP, it shall be SNMPv2 at minimum, using SNMPv3 as a preferred option.
  3. System Access. Access to Network Monitoring System shall be allowed only from internal network segments or through VPN.
  4. User Authentication. Users shall access the System using Diameter, Radius or local authentication, with individual username and password.
  5. User Management. The system shall provide user management capabilities such as creation, modification and deletion of users.
  6. Security Logs. The System shall keep a security log, tracking all security events, including failed access attempts.

Additional security requirements should be considered, as appropriate by the NaaS Operator, complying with any regulatory or customer requirements.

3.5 Additional Features

Up to this point, major requirements for the Network Monitoring System have been analyzed and specified. However, some additional features could be integrated and may be of value for some NaaS Operators.

Some of these requirements are enumerated below for the consideration and evaluation of the NaaS Operators. As previous requirements, they are part of the Functional Requirements Template:

  1. Network Auto-discovery. The System may integrate auto-discovery functionalities to automatically scan the network and add network elements for inventory and monitoring.
  2. Business Level Analytics. The System may provide in addition to network analytics, business-level analytics.
  3. Integration with automated workflows. The System may allow for integration with automated workflows for ticket generation and network automation.
  4. Quality of Experience. The System may support Quality of Experience analysis features.
  5. Protocol & Call Flow Analysis. The System may provide protocol and call flow analysis capabilities in order to detect signaling issues and anomalies in E2E procedures. (Requires packet data capture)
  6. Analysis of log data. The System may support monitoring and analysis of log data from network elements.

4 Implementation Strategy Definition

This Section presents a high-level analysis of implementation options for network monitoring solutions, providing instructions to the NaaS Operator to define its own implementation strategy.

4.1 Topology & Infrastructure

Network Monitoring solutions are integrated by several functional components, as shown in Figure 7. The actual configuration may differ between solutions, but in general, most solutions will preserve the components shown in the Figure.

Figure 7. Network Monitoring Functional Architecture

  • Collection Agents/Packet brokers have the function to pull metrics from devices and third-party APIs, listen for pushed metrics/events, aggregate traffic and manage the data being sent to monitoring tools. A network can have a single or multiple collection agents depending on network size, monitoring protocols and overall requirements.
  • Real-time Processing is the block in charge of rules and event processing and alerting in real-time.
  • Batch processing block applies complex analytics and processing algorithms to transform and extract insights from network data.
  • Long term storage may consist of one or more databases whose purpose is to make data available for queries by the Batch Processing and Dashboard & Reporting.
  • Dashboard and Reporting block is the visualization tool that generates and presents dashboards and metrics.

When evaluating solutions, NaaS Operators should look for these functional components or equivalent, as every vendor or open source project may assign a different name to the specific components of the solution, but functionality should be the same.

Components of network monitoring solutions are usually implemented as software and this should be the preferred option for NaaS Operators as it offers high flexibility and scalability, even if there are some commercial solutions implemented in hardware.

It is important to consider that for certain cases hardware may be a requirement. For example, probes are mainly tied to specific hardware, and when collecting packet data, collectors are most of the time a hardware appliance.

4.1.1 Topology

Topology refers to location and interfaces between network monitoring components. At a high level, 3 models can be specified:

  • Centralized
    With a centralized topology, a single monitoring system instance monitors the whole network. This installation may consist of one or more servers due to hardware limitations, but it is considered centralized as long as all the hardware is installed in the same facility.
  • Distributed
    With a distributed topology, multiple installations of the monitoring system are used to monitor the whole network. Each monitoring system is installed at a different facility and is responsible for monitoring a geographical or administration region / domain.
  • Hierarchical
    The hierarchical topology is similar to the distributed topology, with the addition of a higher-level monitoring instance that requests information from distributed instances to have visibility of the whole network. An example of a hierarchical topology is when the network monitoring system needs to collect data from vendor-specific NMS or OSS.

Table 4 below shows a comparison of these topologies that aims to support the NaaS Operator to define the topology for its own solution.

Criteria

Centralized

Distributed

Hierarchical

Reliability

Single point of failure

Better reliability without a single point of failure. Natural geographic redundancy

Highly reliable. Allows for local and centralized monitoring without a single point of failure

Scalability

Constrained by bandwidth

Scales seamlessly with the network, just by adding a new monitoring instance

Scales locally. The problem of bandwidth can be solved by filtering data that goes directly to centralized monitoring.

Complexity

Easiest to deploy and maintain

Easy to deploy but maintenance requires personnel in various regions

The Integration of distributed and centralized instances can get complex.

Cost

Low

Medium (depends on the number of regions/instances)

High. Basically, it is basically the sum of centralized and distributed

Table 4. Network Monitoring Topology Comparison

NaaS Operator shall consider the criteria above to select an adequate topology. As a general rule, the centralized topology is recommended for most cases. However, there are certain cases in which a hierarchical topology may suit better the NaaS network; these cases are when the NaaS Operator has multiple teams and facilities distributed in various regions and/or when collecting NetFlow and packet data.

A special case of the hierarchical topology is that in which the network monitoring is collecting data from vendors NMS or OSS and not directly from network elements.

It is recommended to implement a monitoring strategy with High-Availability through failover. This characteristic is achieved through redundancy in the network monitoring solution components. The NaaS Operator should consider at least redundancy for power supply, servers and network connectivity.

4.1.2 Network Monitoring Connectivity

Connectivity from the Network Monitoring system towards the network can be achieved with an in-band or out-of-band scheme. The in-band scheme consists of using the network that is being monitored to transport its own monitoring data. On the other side, out-of-band connectivity consists of deploying a secondary network just with monitoring purposes. This is required for most solutions that collect packet data.

For the rural environment of NaaS Operators is highly (and in most cases prohibitive) costly to deploy an overlay network in addition to the RAN and transport network that provides the network services. Thus, the NaaS Operator shall define in-band connectivity for network monitoring. Out-of-band connectivity can be defined only in very specific cases and locations such as transport or core nodes with existing network infrastructure where packet data is being monitored.

4.1.3 Cloud vs On-Premise

For many years and in many networks, network monitoring solutions have been deployed on-premise. However, there is a trend towards cloud-based network monitoring due to lower CapEx investment and fast deployment.

Todays commercial solutions are divided between on-premise and cloud deployment, with some of them offering both options.

The NaaS Operator should evaluate solutions with these two implementation approaches and select the approach that fits better its requirements and constraints. To that extent, Table 5 below makes a comparison between both approaches that can support the NaaS Operator to make the right decision.

Criteria

On-Premise

Cloud

Install & Setup

Installation and setup of on-premise software imply significant effort and time.

Instant setup, complete and pre-configured for rapid deployment with low risk. Devices need to have access to the Internet, which requires an active and secure connection for the whole network.

Scalability

Scaling requires new hardware and infrastructure/connectivity expansion

Highly granular and timely scaling. Expansion of connectivity to the cloud might be required.

Reliability

Owned facilities will have lower reliability than the cloud. However, depending on cloud connectivity the on-prem solution may have better reliability

Highly reliable. A weak point could be connectivity to the cloud.

Cost

Most vendors require the purchase of a perpetual software license; however, some vendors offer subscription models similar to cloud solutions and some open source solutions are license free

Low TCO due to avoidance of upfront costs (hardware, installation, setup) and pay-as-you-grow schemes

Security

High level of data security when set up correctly

Medium to High level of security given that devices are connected to the Internet. It requires diligence to adequately secure the network.

Table 5. On-Premise vs Cloud Comparison.

As a rule of thumb, NaaS Operators with small networks and tight financial constraints may decide for on-premise open source monitoring systems. If not in this case, the Operator should prefer cloud solutions due to setup, maintenance and cost efficiency. Nevertheless, other factors may influence the decision and should be taken into account by the NaaS Operator, such as regulatory constraints, cloud connectivity issues and the pre-existence of on-premise solutions and skills to set up and maintain them within the Organization.

4.1.4 Interfaces to Other Systems

The Network Monitoring System is not isolated in the NaaS environment. Thus, it requires connectivity to and from other systems, such as OSS/BSS, Network Management Systems, Cloud Services, virtualized elements and cloud management platforms.

The connectivity to these systems is increasingly done through web APIs and data models, and the NaaS Operator should specify the support of these interfaces to consume and provide data from/to other systems.

In addition, SNMP interfaces shall be considered as many systems still use this protocol for management connectivity.

Based on the analysis in the sections above, the following recommendations are provided to the NaaS Operator for the definition of Topology & Infrastructure aspects of the Network Monitoring Implementation Strategy:

  • Verify that solutions implement at minimum the functional blocks shown in Figure 6, or equivalent.
  • Prefer software implementations. Consider hardware solutions only when using probes or collecting packet data.
  • The topology should be either centralized or hierarchical depending on the network size, location of personnel and facilities.
  • Consider redundancy at the infrastructure (power and servers) and software instances, cloud & network connectivity.
  • Stick to in-band monitoring unless the overall monitoring strategy considers collection of packet data.
  • To decide between cloud and on-premise deployment, a careful analysis is required considering NaaS Organization monitoring requirements, IT skills, budget and network size.
  • Always consider the support for connectivity with other systems, at least through SNMP and web APIs
  • Verify that solutions implement at minimum the functional blocks shown in Figure 6, or equivalent.

  • Prefer software implementations. Consider hardware solutions only when using probes or collecting packet data.
  • The topology should be either centralized or hierarchical depending on the network size, location of personnel and facilities.
  • Consider redundancy at the infrastructure (power and servers) and software instances, cloud & network connectivity.
  • Stick to in-band monitoring unless the overall monitoring strategy considers collection of packet data.
  • To decide between cloud and on-premise deployment, a careful analysis is required considering NaaS Organization monitoring requirements, IT skills, budget and network size.
  • Always consider the support for connectivity with other systems, at least through SNMP and web APIs

4.2 Scalability

Implementation of the Network Monitoring System must consider scalability requirements to ensure that it will support current and future load in terms of data collection, processing and storage.

The NaaS Operator may use the following guidelines to define scalability requirements:

  • Network Element and Service Types. Validate that the Network Monitoring System supports monitoring of all the network element types in the network as well as service monitoring through the provided APIs.
  • Monitoring Bandwidth & Processing. Validate that the Network Monitoring System has enough bandwidth and processing capacity to cope with the traffic volume and frequency generated by the current and forecasted number of monitored elements and the corresponding collection frequency.
  • Storage. Validate that the Network Monitoring System incorporates enough storage capacity for the ingestion and long-term storage of network data for a defined period of time that shall be one year at minimum.

4.3 Procurement Strategy

Acquisition and implementation of the Network Monitoring System can be done in various ways, depending on the size of the organization, skills and financial constraints. At a high-level three aspects must be considered by the NaaS Operator

  1. Open source vs Proprietary Software
  2. Insource vs Outsource implementation
  3. Single vendor vs Best-in-breed

These aspects are discussed in the following sections.

4.3.1 Opensource vs Proprietary Software

Opensource refers to free or low-cost software that allows to see and customize its code. The developer community is responsible for the performance and improvement of the tool. The NaaS Operator may customize and improve the code to meet its needs, which may require significant time and effort depending on the level of customization desired.

Support for opensource solutions is offered through forums and articles. For certain tools, there exist robust communities that provide all the required support.

Regarding cost, many opensource platforms provide the code free of charge, however, there are times when an opensource monitoring solution requires users to pay for extra features and added functionalities.

On the other hand, proprietary software will have a cost, with varying pricing schemes which can be based on the number of network elements, bandwidth, storage capacity or a one-time fee. In this case, performance is ensured by the vendor, which offers proven functionality that is constantly improved, and it generally includes the option to contact tech support agents.

The main drawback (apart from cost) for proprietary software is the lack of customization. Even if the tool allows for customization of dashboards, reports, etc. it is tied to the existing code. Specific features required by the NaaS Operator may never see the light of the day as the tool follows each vendors roadmap. In this case, its important that if choosing proprietary software, the NaaS Operator makes sure that the tool has all the required functionalities.

The final decision depends on how the NaaS Operator weights each criterion. If cost or flexibility is the main driver, the best option is opensource. However, if performance is critical and there is an important need for support, the NaaS Operator should select proprietary software.

In addition, when choosing opensource software, the NaaS Operator must check any relevant legal issues and compliance requirements to make sure the software can be used for commercial use.

The NaaS Operator can get started taking a look at the Monitoring Tools in the Tool Catalogue provided.

4.3.2 In-source vs 3rd Party Implementation

This aspect refers to the resources that will ultimately implement the solution: personnel from the NaaS Operator or a 3rd Party company which can be either the software vendor or a system integrator.

For the NaaS Operator this should be a matter of weighting availability and skills of the internal resources vs required timelines for the implementation. If these resources are not available or they do not have the required skills and expertise to guarantee an effective and efficient implementation, then the NaaS Operator should prefer a 3rd Party Implementation.

4.3.3 Single vendor vs Best-in-Breed

Finally, the NaaS Operator should define whether it prefers to implement the complete Network Monitoring System through a single vendor or to apply a best-in-breed approach. The best-in-breed approach allows to select the best option for each functional block of the solution: hardware/cloud provider, collection agents, storage, processing and dashboards.

The best-in-breed applies mainly to opensource implementations, but it can also be applied to proprietary software, where the NaaS Operator or a System Integrator integrates a solution through components from various vendors.

The recommendation is for NaaS Operators that are just getting started to go with a single vendor or single opensource project. Then as the network matures or if the NaaS Operator has previous expertise, start incorporating various vendors or opensource projects.

5 Network Monitoring: Reference Architecture & Requirements Specification

After going through the previous sections, the NaaS Operator has been enabled to establish its monitoring strategy, specify functional requirements for the network monitoring solution and define the implementation strategy.

Now, the NaaS Operator is encouraged to follow the procedure below and use the provided Network Monitoring Architecture Template in order to document these decisions and communicate requirements to the overall NaaS Organization and potential vendors.