改进蚁群路由算法在多媒体CDN中的应用研究

Application Research of Improved Ant Colony Routing Algorithm in Multimedia CDN

多媒体内容分发网络是将源服务器(Origin Server)中的多媒体内容分散到多个边缘服务器(Replica Server)上,以降低服务器的负载并提升系统的可用性﹒文章提出以改进蚁群路由算法(AntColony-based Routing Algorithm,ARA)来解决这一问题﹒首先利用ACO(Ant Colony Optimization)算法将多媒体内容发送到各边缘服务器,边缘服务器再回送其状态信息至源服务器;然后,源服务器更新请求路由系统中的内容数据库,使用费洛蒙函数让蚂蚁在每个决策中选择下一个节点﹒仿真结果表明,ARA算法不仅具有负载平衡能力,还可增加重定向命中率,且记录时间超1 000 s时,ARA算法比RIP算法重定向命中率要高2%～8%﹒

The multimedia content distribution network distributes the multimedia content in the origin server to multiple Replica Server to reduce server load and to improve the system availability. An Ant Colony-based Routing Algorithm（ARA） to solve this problem is proposed in this paper. First, the ACO algorithm is used to send the multimedia content to each edge server, and the edge server sends back its status information to the source server, then the source server updates the content database in the request routing system, and uses the pheromone function to let the ants choose the next node in each decision. The simulation results show that the ARA algorithm not only has load balancing ability, but also can increase the redirect hit rate. When the recording time exceeds 1 000 s, the ARA algorithm is 2% to 8% higher than the RIP algorithm.