基于遗传算法的Peer-to-Peer路由算法R-GA

路由算法是制约Peer-to-Peer 系统整体性能的关键因素之一.目前大多数路由算法无法保证全局收敛,而链路延迟、费用、网络带宽等现实制约因素往往在选路时被忽略.针对上述问题,提出了基于遗传算法的R-GA路由算法.通过适度函数和遗传因子,R-GA可以快速地实现全局收敛.同时将链路的延迟、费用、带宽等参数插入到适度函数中, 避免了盲目路由.仿真试验的结果表明,R-GA路由算法在大规模Peer-to-Peer系统中是高效和可扩展的.

Novel Routing Algorithm R-GA in Peer-to-Peer System Based on Genetic Algorithms

The routing algorithm is one of the most crucial factors which can restrict the performance of the whole Peer-to-Peer systems.The optimal solutions of most existing distributed routing algorithms have the local convergence problem,and take no consideration for the practical conditions,such as the latency,cost and bandwidth.To address these problems,a novel routing algorithms R-GA is proposed,which is based on Genetic Algorithm(GA).By using the fitness function and the genetic operators,(R-GA) can get a global convergence solution rapidly.In order to avoid the blind routing,the latency,cost and bandwidth parameters are inserted into the fitness function.The simulation results show that the RGA routing algorithm is efficient and scalable in the large-scale Peer-to-Peer systems.