一种基于QoS的选播路由算法研究  

Research on Adaptive Anycast Routing Algorithm with Multiple QoS Parameters

作　　者：李陶深[1,2] 陈松乔[1] 陈建二[1] 陈燕[2] 冯凌凌[2] 

机构地区：[1]中南大学信息科学与工程学院,长沙410083 [2]广西大学计算机与电子信息学院,南宁530004

出　　处：《计算机研究与发展》2008年第z1期415-420,共6页Journal of Computer Research and Development

基　　金：广西壮族自治区自然科学基金项目(06400026);广西“新世纪十百千人才工程”专项基金项目(2001213号)

摘　　要：引入适应度尺度变换,提出一个基于QoS的选播通信服务模型和选播路由算法.算法首先根据选播组的大小把整个网络图的路径搜索空间动态地划分为k个子空间,每个子空间仅包含一个该组成员;其次对每个子种群进行初始化和编码;然后在几个既不重叠、又能反映整体性质的子空间上并行搜索和演化,直至求得满足多个QoS参数约束的选播路由最优解.网络仿真模拟实验结果表明该算法是有效且切实可行的,它能满足多个QoS的约束条件,较好地平衡了网络负载,改善了网络服务质量.By introducing the concept of fitness scaling, a anycast service model with multiple QoS parameters constrained and a new anycast multi-path routing algorithm are proposed. The idea of this algorithm is as follows:1) dividing dynamically path search space of the whole network diagram into k sub-space according to the size of anycast group, and each sub-space only contains one member of this group; 2) initiating and coding per sub-population; and 3) parallel-searching and evolving on the several non-overlapping sub-space until the optimum solution of anycast routing with multiple QoS parameters constrained is obtained. The results of network experiment show that this algorithm is effective and available. It can satisfy the constrained condition of multiple QoS,balance network load fairly, and improve the quality of network service.

关 键 词：选播路由 QOS 自适应遗传算法 通信服务模型 多QoS参数约束 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]