基于节点信誉改进Q学习的P2P资源搜索策略  

Improved resource search strategy using Q-learning based on node reputation for P2P network

作　　者：刘焕淋[1] 陈高翔[1] 吴帅勇[1] 

机构地区：[1]重庆邮电大学通信与信息工程学院,重庆400065

出　　处：《重庆邮电大学学报（自然科学版）》2013年第6期801-806,共6页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：国家自然科学基金(61275077;61371096);重庆市自然科学基金(cstc2013jcyjA40052)~~

摘　　要：针对对等网络(peer to peer,P2P)中资源搜索效率低的问题,提出一种基于节点信誉改进Q学习的资源搜索策略(search strategies using improved Q-learning based on node reputation,SSQBR)。该策略在Q学习模型基础上通过引入信誉机制,在搜索初期通过选择信誉值最高的邻居节点转发查询请求以解决Q表的资源信息不足时搜索成功率低的问题,该策略加速了Q学习过程,提高了资源搜索的成功率。仿真结果表明,改进的资源搜索策略与其他搜索策略法相比,可有效降低资源搜索响应时间,提高搜索成功率。Aiming at the low efficiency of resource search in P2P network, a search strategy using improved Q-learning based on node reputation （SSQBR）is proposed in the paper. By introducing a reputation mechanism based on the Q-learn- ing model, the proposed SSQBR strategy can select the node with maximum reputation value to forward the query request at the beginning of search process, which can solve the problem of low resource search success rate when the Q table informa- tion is not sufficient. The SSQBR strategy accelerates the process of Q-learning and improves the resource search efficiency. Comparing with other resource search strategies, the simulation results show that the proposed strategy can reduce resource search response time and increase the search success rate.

关 键 词：对等网络(P2P) 资源搜索 Q学习 节点信誉 成功率 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]