An Optimal Route Computation Method for GMPLS Based Survivable Network with a Tabu Search Algorithm for the Weighted Constraint Satisfaction Problem

This paper proposes a route computation method for a GMPLS based survivable network. It is classified as a precomputed and centralized approach. It tries to globally optimize both working and recovery routes for each demand while taking account of GMPLS recovery schemes and network resource constraints. To enhance its scalability as well as computational performance, two unique devices are introduced to the method: utilization of a tabu-search algorithm for the weighted constraint satisfaction problem (WCSP) and the limitation of candidate routes. Experimental results demonstrate the possibility that it could handle practical-scale networks with satisfactory performance in both optimality of solution and computational time