基于惩罚因子的多约束剪枝QoS路由算法  

Multiple Constraint Pruning Qo S Routing Algorithm Based on Penalty Factor

作　　者：赵季红[1,2] 张方[3] 曲桦[1] 戴慧珺[1] 

机构地区：[1]西安交通大学电子与信息工程学院,西安710049 [2]西安邮电大学通信工程系,西安710061 [3]西安交通大学软件学院,西安710049

出　　处：《计算机工程》2015年第4期70-76,共7页Computer Engineering

基　　金：国家自然科学基金资助项目(61371087);国家科技重大专项基金资助项目(2013ZX03002010-003;2012ZX03001023-003)

摘　　要：为满足新一代互联网上各业务的服务质量(Qo S)要求,设计一种多约束剪枝Qo S路由算法PF-MCP,采用剪枝算法思想,综合考虑带宽、时延和抖动多重Qo S约束,引入惩罚因子进行综合评价找出多条有效路径并从中选取最优路径。采用固定和随机网络层拓扑进行仿真实验,结果表明,与合成优先权值路由算法相比,PF-MCP算法在Qo S满意度、服务节点计算能力及时间复杂度方面性能较优,具有路由运算量小、网络恢复能力强及选取的最优路径可信性高的特点,并且在覆盖网Qo S路由方面适用范围广。In order to fulfill the various Quality of Service（ Qo S） requirements of new services on the next generation Internet,this paper proposes a novel overlay Qo S routing algorithm,w hich is named Penalty Factors-M ultiple Constraint Pruning（ PF-M CP） algorithm. The idea of pruning is applied by PF-M CP to handle multiple Qo S constraint,such as bandw idth,delay and jitter. In order to select the optimal solution,penalty factor is introduced by PF-M CP to comprehensively evaluate the effective paths. Simulation experimental results by using fixed and random netw ork layer topology show s that PF-M CP algorithm has higher Qo S satisfaction rate,service node computation capability equilibrium and time complexity compared w ith Synthesize Priority Weight Routing（ SPWR） algorithm,it has characteristics of small amount of routing calculation,strong netw ork recovery capability and the optimal path selecting is credible. In summary,PFM CP can effectively fulfill its overlay Qo S routing.

关 键 词：覆盖网 服务质量路由 多约束 剪枝算法 惩罚因子 网络拓扑 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]