Optimized network planning for converged optical access networks
Security and Dependability, Operating-, Communication- and Distributed Systems
Final Report Abstract
Growing user requirements and the need to address the new services and user groups, i.e., Machine-to-Machine in 5G, as well as high competition force network operators to look for effective network planning solutions. The goals of the planning can be widespread from reducing energy consumption to proving the required reliability. In this project, we holistically investigated the strategic network planning from the input preparation to the network planning methods and techno-economic evaluation. The results of the theoretical investigations were verified on the real geographical data and can be reproduced or extended with a publically available geography-based fixed network planning tool. The results of this project can be directly economically applied not only to the telecommunication planning, but to all types of infrastructure networks, e.g., power or water. Strategic network planning relies on the qualitative input data: technology characteristics, demand models and models of the underlying geographic area. We compiled a database of passive optical network characteristics relevant for planning: technical and costs. Further, we looked into learning non-Poisson distributed traffic from its traces to model the traffic demands. We concluded the input modeling by looking into the influence of the input geographic area on the planning results, comparing the planning outcomes for the reference scenario of the geographic map to a benchmark graph model and a proposed procedurally generated graph topology. This comparability and reproducibility was achieved by creating a geography-based fixed network planning tool. This tool is publicly available, free to use, extendable and allows using the developed planning methodology. The state-of-the-art planning methodology had to be updated for converged network planning to study the synergy effect from including heterogeneous demands in terms of reducing the overall network resources consumption, reducing the cost of the network implementation and exploring the benefits from the existing technologies that would not be available in the separate networks. We investigated how the network planning can increase possible energy savings, increase network reliability. We have compared our heuristic performance with the proposed in the project optimization formulation (Mixed Integer Program), which can be used as a planning benchmarking scenario providing the upper bound for the network planning quality (lower bound for the costs). The network planning methodology was evaluated on realistic 5G examples of converged access and x-haul networks: residential access and Intelligent Transportation Systems backhaul; or residential access and 5G x-haul. We learned that for the converged access and x-haul networks, the most relevant virtualization application is Cloud Radio Access Network (CRAN) function split as it directly influences the requirements on the x-haul. We investigated its influence on the converged planning for greenfield and brownfield scenarios in urban and rural areas. Finally, we looked into the protection options for RAN and converged optical distribution network (residential access and 5G RAN). The network planning tool is publically available at: https://github.com/EGrigoreva/FixedNetworkPlanningTool
Publications
- M2M Wake-ups over Cellular Networks: Over-the-Top SIP. Workshop on All Things Cellular in conjunction with ACM MobiCom 2016
E. Grigoreva; J. Xu, W. Kellerer
(See online at https://doi.org/10.1145/2980055.2980059) - Energy Consumption and Reliability Performance of Survivable Passive Optical Converged Networks: Public ITS Case Study. Journal of Optical Communications and Networking (JOCN) Volume: 9, Issue: 4, 2017, C98 - C108
E. Grigoreva, E. Wong, M. Furdek, L. Wosinska, C. Mas Machuca
(See online at https://doi.org/10.1364/JOCN.9.000C98) - Heterogeneous Wireless Access Network Protection for Ultra-Reliable Communications. IEEE Vehicular Networking Conference (VNC), 2017
E. Grigoreva, D. Shrivastava, J. Dittrich, H. Wilk, H.-M. Zimmermann, C. Mas Machuca, W. Kellerer
(See online at https://doi.org/10.1109/VNC.2017.8275612)
