A Fuzzy Genetic System for Resilient Routing in Uncertain & Dynamic Telecommunication Networks

Network connectivity has become an essential part of our modern society, so it is important for telecoms organizations to be able to provide resilience to network faults. A solution to this is to provide multiple distinct routes between network locations. Furthermore, network conditions can fluctuate which in turn introduces high uncertainty levels. Due to the nature of this problem traditional greedy routing methods such as Dijkstra's algorithm and A* struggle to provide an optimal solution, as they will use the best route through a network for the first route and leave a more expensive route for a secondary route.In this paper, we will present a fuzzy genetic system that can provide multiple routes between two locations that is robust to uncertain environments. The proposed system is based on the combination of Dijkstra's Algorithm, Genetic Algorithm and Type 1 Fuzzy Logic. We have tested the proposed system with real-world telecoms infrastructure data, which is affected by high levels of uncertainty. The results showed that the proposed system can generate more consistent routes in 30% more cases and finding consistent routes in 17.5% more cases when compared to Dijkstra's Algorithm. The fuzzy implementation of the proposed system increased the number of consistent routes by a further 14.90%.