An Improved Genetic Algorithm for Cost-Delay-Jitter QoS Multicast Routing

In this paper, an improved genetic algorithm is presented to solve the multicast routing problem, which is known as NP-complete problem. The contribution of this work includes: (1) the adaptive niche technique and new migration rules are designed to improve the performance of genetic algorithm. (2) In order to fulfil the adaptability of the basic operators of genetic algorithm, the artificial immune system is involved to dynamically control the crossover operator and the mutation operator. (3) A tree encoding based on the theory of generating the spanning tree is proposed to map the solution space of the multicast tree of QoS multicast routing problem, which basic definitions and axioms of the topology are described especially the character of spanning tree and the nature of the cut edge and cut set of a tree. Experiment result shows that this improved genetic algorithm has higher accuracy and performance than traditional methods