摘要
针对遗传算法强大的全局寻优和并行处理能力 ,提出了一种改进的遗传算法调度ATM网络动态路由的新算法。综述了使用遗传算法进行通信网络节点选择的问题 ,指出了原有算法进行遗传算法优化时存在的不足 ,并提出一种新的基于动态惩罚的解决策略 ,其思想是在进化的过程中 ,逐渐减小对可行解惩罚的力度 ,使网络向可行解逼近。通过简单模型的仿真 ,验证了动态惩罚策略的可行性和有效性。
An improved routing algorithm in ATM network based on genetic algorithm is proposed because GA is employed to solve routing problem for its strong optimization and parallel computing ability. The paper surveys the faults of network optimization using old algorithm, and a new method based on dynamic penalty is presented. This method employs the varying weight coefficient of penalty while GA process is proceeding. The simulation result shows that the operation is feasible and valid.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2003年第2期216-218,共3页
Systems Engineering and Electronics
关键词
ATM网络
路由选择
遗传算法
动态惩罚
ATM network
Routing selecting
Genetic algorithm
Dynamic penalty
