摘要
IP/DWDM光Internet是下一代互联网主要组网技术之一,服务质量组播路由算法是其一个研究热点,已经证明它是NP完全的。为此设计了一种非NP类算法,基于尺度收缩混沌变异进化算法,构造一棵带宽、延迟、延迟抖动和出错率受限且费用优化的QoS组播路由树,兼顾网络负载均衡。仿真研究表明,该算法是可行和有效的,较好地克服了传统遗传算法的早熟和收敛性差的问题。
IP/DWDM optical internet is one of the main networking technologies for the NGI （next generation internet）. QoS mUltlCaSt routing algorithm is a research focus which is proven to be NP-complete. Thus, a non-NP algorithm is proposed, It is based on the shrinking-chaotic-mutation evolutionary algorithm to find a bandwidth, delay, delay jitter and error rate bounded and cost optimized QoS multicast routing tree. The network load balancing is taken into account at the same time. Simulation results show that the proposed algorithm is both feasible and effective with the improved search ability and convergence speed to the optimal solution over its counterpart based on the traditional genetic algorithm.
出处
《计算机工程与设计》
CSCD
北大核心
2006年第17期3114-3116,3122,共4页
Computer Engineering and Design
基金
国家自然科学基金项目(60473089
60003006
70101006)
辽宁省自然科学基金项目(20032018
20032019)
中国教育科研网格ChinaGrid
教育部现代远程教育工程项目(200403)。
