The Research Based on Dual Population Genetic Algorithm for Traffic Engineering in MPLS Networks
Cai Qun-ying · Jisuanji fangzhen · 2010
Rapid growth of Internet traffic and increasing requirements for service quality have made traffic engineering an essential consideration in the operation of large internet backbone network.This paper introduces the traffic engineering working-mechanism and math model.The key problem of the MPLS TE is the optimized distribution of LSPs.As to the problem's character,this paper introduces a genetic algorithm with two populations to solve it,and it uses the natural number coding method to improve search efficiency.In this algorithm,two subpopulations are evolved independtly and exchange best chromosome in a periodic way.Simulation shows that the algorithm can effectively speed up the searching process and boost the chance of achieving the best results.