Rapidio网络QoS多目标优化  

Multi-objective optimization of Rapidio network

作　　者：蔡炜[1,2] 张建东[1,2] 蔡惠智[1,2] 

机构地区：[1]中国科学院声学研究所,北京100190 [2]北京中科海讯电子科技有限公司,北京100107

出　　处：《计算机应用》2010年第A12期3172-3175,共4页journal of Computer Applications

摘　　要：针对现有Rapidio网络路由管理策略的缺点,提出了一种基于改进的遗传算法的优化策略。通过改进的路由分配策略以及对传统遗传算法编码、交叉和变异等方面的改进,明显提高了Rapidio网络服务质量(QoS)诸项指标;同时相对于传统遗传算法,改进的遗传算法拥有更快的收敛速度。该算法适用于基于Rapidio网络的嵌入式应用,具有较强的工程实用价值。Concerning the drawback of the existing Rapidio network routing management strategy,an optimized strategy based on improved Genetic Algorithm(GA) was proposed.By doing some improvements on the routing strategy,coding,mutation and crossover of conventional GA,the performance of the Quality of Service(QoS) objects was improved effectively.At the meantime,compared to the conventional GA,the improved GA increases convergence speed greatly.The proposed algorithm is suitable for embedded applications based on Rapidio network,and has strong application value.

关 键 词：Rapidio网络 遗传算法 服务质量 适度函数 算法编码 

分 类 号：TP368.12[自动化与计算机技术—计算机系统结构]