Using memory to improve the VNS metaheuristic for the design of SDH/WDM networks

Abstract

Variable neigborhood search is among the well studied local search based metaheuristics. It has provided good results for many combinatorial optimization problems throughout the last decade. Based on previous successful applications of this metaheuristic on various network design problems in telecommunications, we further enhance this approach by incorporating adaptive memory mechanisms from the scatter search and tabu search metaheuristics. The heuristics are compared among each other as well as against objective function values obtained from a mathematical programming formulation based on a commercial solver. The problem instances cover a large variety of networks and demand patterns. The analysis carried out in this paper corroborates that there are significant differences between the variable neighborhood search and the hybrid approach. © Springer-Verlag Berlin Heidelberg 2006.