用户行为统计特性对通信网络性能可靠性的影响

摘要

针对二维方格网络模型，对用户访问网络资源过程中呈现的幂律特性和集聚特性进行了建模，借助网络相变分析了用户访问行为的统计特性对网络性能可靠性带来的影响。研究结果表明，用户访问资源节点形成的二分网络的度分布的幂律特性对网络性能可靠性的影响较小，但这种二分网络中具有高访问量的资源节点子集的频繁衍变会改善网络的性能可靠性；用户访问行为的集聚特性中突发性访问的强度、频率和持续时间在一定程度上影响着网络性能可靠性的相变行为；随着突发性访问与正常性访问时间单元之间活动用户数量比值的增大，网络性能可靠性的相变点逐渐右移，即网络更加能够容忍高强度的突发性访问。%Based on a lattice network, a user access behavior model was established to describe the statistical characteris-tics of user behavior including the clustering characteristics and the power law distribution. The phase transition point was used to analysis the influence of the statistical characteristics of user behavior on the network performance reliability. It was shown that the network performance reliability was almost independent on the degree distribution of the serv-er-visited bipartite network; however the hot changing of the small subset with the high degree server nodes improved the network reliability. The strength, frequency and the duration of the clustering characteristics have certain impact on the phase transition of network performance reliability. The phase transition point of network performance reliability moved to the right with the increase of the difference between the number of active users when the clustering characteristics happened and don’t.