Abstract: Based on genetic algorithm and the extremal optimization idea, this paper propose an extremal genetic algorithm to deal with the QoS multicast routing. The non-equilibrium of extremal optimization can avoid to local minimums and accelerate convergence. According to the characteristics of the networks, a suit of special encode, crossover and mutation is used to guarantee the feasibility of the solution. Simulation results indicate that this algorithm has better performances on both speed and precision.

Key words: QoS multicast routing, multicast tree, extremal optimization, genetic algorithm

摘要： 基于遗传算法和极值优化思想，提出一种极值遗传算法，将其应用到QoS组播路由。极值优化的非自衡性可以防止算法陷入局部解，加快算法的收敛速度。根据网络拓扑结构特点，采用特殊的编码、交叉、变异操作，保证解的可行性。实验结果表明，该算法能达到较高的QoS组播路由速度和精度。

关键词: QoS组播路由, 组播树, 极值优化, 遗传算法

CLC Number: 

• TP393