摘要
为了避免传统启发式算法在求解多播路由问题时存在的过早收敛问题,提出了一个新的动态多播路由免疫算法(DCOM IA),此算法利用克隆选择和基因库的思想改善了群体的多样性,并评估了二进制串表示的候选个体.同时,提出了一个改进了的动态约束多播路由问题(M DCM R),试验结果表明:此算法求解该动态多播问题是高效的.
In order to avoid the too-early-convergence phenomenon existed in traditional heuristic algorithm for multicast routing problems, a new random algorithm, namely Dynamic Constraint On-line Multicast Immune Algorithm (DCOMIA) is proposed. DCOMIA uses clone selection and gene library to improve the population diversity and evaluation of candidate solutions which are represented by binary strings. Moreover, a revised dynamic constraints multicast routing problem, namely MDCMR, is defined and the initial experiments on MDCMR problems shows that DCOMIA is very promising to improve the performance of dynamic multicast routing.
出处
《中南民族大学学报(自然科学版)》
CAS
2005年第4期64-67,共4页
Journal of South-Central University for Nationalities:Natural Science Edition
基金
国家自然科学基金资助项目(6007304370071042)
关键词
动态多播路由
免疫算法
基因库
dynamic multicast routing
immune algorithm
gene library