对等网络中改进蚁群智能搜索算法研究

Intelligent Search Study Based on Improved Ant Colony Algorithm in P2P Networks

摘要: 为了提高蚁群算法在P2P网络资源搜索中存在搜索盲目、搜索效率低的问题，论文将多态蚁群算法和应用到了P2P网络搜索。针对搜索一段时间后网络中发起的对新的文件请求，引入合成信息素的概念，以减少搜索初始阶段消息转发的盲目性。对无结构P2P网络中的洪泛算法、蚁群算法、引入合成信息素后的蚁群算法进行模拟实验，实验结果表明所提出的算法可有效提高P2P网络的搜索性能。

Abstract: In order to enhance the practicality of ant colony algorithm and improve the search efficiency of peer-to-peer networks, this paper presents a new approach of unstructured P2P information re-trieval based on the polymorphic ant colony algorithm. In order to meet the new file requirement after a while of searching, the conception of generated pheromone is imported to decrease the blindness of pack forwarding in early searching stage. Based on the simulator framework, simu-lating the flooding, ant colony algorithm, ant colony algorithm with generated pheromone in un-structured peer-to-peer networks, and analyzing the experience data, the experience results indi-cate that the algorithm is effective and can enhance the performance of peer-to-peer networks.