小生境粒子群优化ABC支持型QoS组播路由机制

ABC supporting QoS multicast routing mechanism using niche particle swarm optimization

针对ABC总最佳链接支持型QoS(服务质量)组播路由存在的链路参数不精确、用户QoS参数难以描述以及用户与网络运营商效用相矛盾等关键问题,提出了一种基于用户和运营商博弈的多目标模型,采用模糊数学的方法对链路参数和用户的QoS需求进行处理,利用用户和运营商在每条链路上效用的博弈达到纳什均衡下的Pareto最优,保证了用户和运营商的公平性.为有效求解该多约束QoS组播路由模型,提出一种基于动态Pareto解聚类分析的小生境粒子群的多目标算法.该算法采用聚类小生境保证解的多样性,引入动态更新Pareto最优边界加速寻得优质解过程.最后,基于NS2平台对路由机制及算法进行了仿真实现与性能指标评价.实验结果表明:该路由算法在可信度、用户效用、网络运营商效用等方面具有突出的优势,验证了其有效性与可行性.

To solve these problems that the ABC（always best connected）supporting QoS（quality of service）multicast routing mechanisms usually suffer from the problems of inaccurate link parameters,indescribable QoS parameters of user and the conflicting interests between users and operators,a multi-objective model based on user and operator of the game was proposed.Specifically,this model characterized those indescribable parameters by using the fuzzy math theory and ensured the fairness between users and operators by using the theory of Nash equilibrium.To solve this multi-constrained QoS multicast routing model effectively,a multi-objective optimization algorithm,namely niche particle swarm optimization based on dynamic Pareto cluster analysis,was developed.This algorithm employed clustering niche to keep diversity of the population,and received high-quality solutions as soon as possible by introducing the dynamic Pareto analysis.Finally,a comprehensive simulation experiment was conducted on the NS2（Network Simulator 2）platform.Experimental results show that the proposed mechanism obtains satisfactory performance regarding credibility of path,the utilities of users and network providers,which explicitly verifies its effectiveness and feasibility.