基于蜂群的ABC支持型QoS单播路由机制

ABC Supporting QoS Unicast Routing Scheme Based on Beehive Algorithm

下一代互联网NGI(Next-Generation Internet)需要提供服务质量QoS(Quality of Service)路由,支持总最佳连接ABC(Always Best Connected).但是,由于链路状态的难以精确测量与用户QoS需求的难以准确表达,因此路由机制应该具备处理模糊信息的能力.同时,在网络运营日益商业化的环境下,路由时应该兼顾网络提供方与用户利益,支持各方效用共赢.为此,本文引入模糊数学和微观经济学有关知识,设计了一种ABC支持型QoS单播路由机制.该机制采用区间描述用户柔性QoS需求,使用边适合隶属函数适应链路状态不精确,引入边带宽定价、边评判和路径评价,基于蜂群算法,寻找使用户与网络提供方效用达到或接近Nash均衡下Pareto最优的QoS单播路径.仿真结果表明,该机制是可行和有效的.

NGI （Next Generation Intemet） needs to provide QoS （Quality of Service） routing and support ABC （Always Best Con- nected）. However,due to the difficulty on the exact measurement of link status and the exact expression of the user QoS requirement, QoS routing should be capable of dealing with fuzzy information. Meanwhile, with the gradual commercialization of the network opera- tion, both the network provider and the user profits should be considered and thus their utility win-win should be supported when rou- ting. In this paper,by introducing the knowledge of the fuzzy mathematics and the microeconomics,a QoS unicast routing scheme with ABC supported is proposed. The range is used to describe the flexible user QoS requirement and the edge suitability membership func- tion is adopted to handle the inaccurate link status information. With the edge bandwidth pricing, edge evaluation and path evaluation introduced and based on the beehive algorithm, it tries to find a QoS unicast path with Pareto optimum under Nash equilibrium on both the user utility and the network provider utility achieved or approached. Simulation results have shown that the proposed scheme is both feasible and effective.