by TIP News
Background
We are using wireless technology more and more, however, pretty soon after the signals have left the radio interface and the communication is about to traverse the network, optical transmission technology is used. To bridge longer distances (roughly more than 100 km) or to use the optical fibres more efficiently with multiple wavelengths on the same fibre, optical DWDM transmission systems are used. These optical systems have been developed to transport as efficiently as possible as many bits per second as far as possible. A key solution for this is the optical amplification of multiple connections in one device. The optical systems are planned for best performance vs cost. Once an optical line system (OLS) has been designed, wavelength connections can be planned by estimating the signal quality expected at end of life (EoL). Here we have two important steps, network design and connection planning. Optical systems started as supplier specific, closed solutions which fostered the supplier unique designs. Now the industry is deploying disaggregated optical systems designed with equipment from several suppliers.
Why a supplier independent planning tool for optical networks?
The telco operators are competing to provide the most attractive services and are therefore continuously reviewing the costs for their solutions including the optical systems. Comparing different offers in terms of price was relatively simple, but comparing the performance of closed optical systems from different suppliers has been a challenge. Each supplier has developed their own design metrics and tools based on different approaches and margins, so benchmarking different suppliers’ systems becomes hard. For the disaggregated optical systems, another challenge occurs: How to plan networks based on a mix of equipment from different suppliers?
Until TIP OOPT proposed the GNPy (Gaussian Noise Model in Python) no open-source supplier-independent framework for optical network design and performance estimation of connection was available. For many operators today, GNPy provides the basis for design and connection verification of closed, open, partial and fully disaggregated optical systems.
Challenges
To be able to perform benchmarking and connection verification it is important that the input data is based on the same definitions, for example network element specification for amplifiers from different suppliers shall be comparable. In the open disaggregated network scenario where a mix of supplier equipment are used, a reference definition must be agreed for consistent analysis of performance.
In the past the operators had to rely on the suppliers specific planning and design tools. For the open disaggregated optical line systems, the operators can use planning and design tools provided by an OT supplier or the O-OLS supplier. In both cases, suppliers would have to adapt to the selected supplier’s specific definition of describing data. However, this would generate a new framework of definitions for the operator every time a new O-OLS supplier’s equipment is being deployed.
This problem disappeared with GNPy that is based on generic descriptions and specifications. Operators can now easily perform benchmarking of performance in RFPs and carry out connection verification in deployed multi-supplier networks like the open disaggregated optical networks.
For a partially disaggregated network, the performance for individual supplier equipment (O-OT and O-OLS) is regulated under NDA which has to be respected when planning with GNPy. In the society today where security becomes more important, and the operator environments become regulated by national security guidelines the support for privacy is fundamental. Therefore the GNPy simulations, commonly defined equipment models, can be shared either in a public database or in a private database only accessible by the operators.
Optical system suppliers have invested a lot in preparing tools for accurate network design and impairment validation but GNPy offers something new, an open framework that gives the suppliers the possibility on a common base to share and show their offers to operators. Suppliers can share their best cost or performance optimised solutions (cost/performance) in a way that is verifiable by operators to appreciate the true value of performance gains.
Summary
GNPy offers an open-source framework for design and impairment validation instead of the proprietary solutions that each optical system supplier provides. When building an on-line SDN architecture and integrating the GNPy we will have an accurate and reliable solution for connection planning. In TIP OOPT the Physical Simulation Environment working group caters for the development of the GNPy. This working group consists of academia, research institutes, operators and suppliers. The acceptance of the GNPy is a crucial contribution to industry. More stakeholders are welcome to join the PSE working group to work on a truly vendor-independent implementation and mature open and disaggregated optical networks.
More about GNPy: