A Two-Stage Genetic Algorithm for the Multi-multicast Routing

摘要

The multi-multicast routing problem is to find a set of multicasts which allows transmission and routing of packets from many sources to many destinations. Due to the amount of residual bandwidth of a link, the problem becomes complicated. This paper presents an approach to solve the problem based on a two-stage genetic algorithm. The problem is decomposed into two objectives, the primary objective is to find a multicast group from a source to its destinations satisfied bandwidth and delay constraints, and the secondary objective is to find an optimal combinatorial solution in all multicasts. For the objectives, two evolving modules are designed. In the first module a method of encoding tree-like chromosome, and corresponding crossover and mutation operator are presented. The numerical simulation shows that the algorithm is efficient to solve the multi-multicast routing problem.