基于大型复杂网络拓扑构建优化算法研究

Research of Topology Optimization Algorithms Based on Large and Complex Network

传统的网络拓扑构建算法在运用到大型复杂网络中时,大多采用启发式算法,节点容量参数由前面的经验值确定,一旦经验值规定错误,会给后面的拓扑结构构建带来干扰,造成启动延迟、服务器负载、消息开销和抖动率等问题。为此提出一种有校验小区域参数更正的复杂网络拓扑构建算法,把复杂网络分成几个不同的小区域,在不同的区域中,对前面反馈的参数进行校验,对经验值参数进行核对,运用纠错方法,保证参数设置的合理性。实验证明新提出的算法相对于传统方法在启动延迟、服务器负载、消息开销和抖动率等参数上,有显著的性能提高。

Traditional network topology algorithm applied to large complex networks. Put forward a kind of small area pa-rameter correction of the complex network topology construction algorithm, the complex network is divided into several small areas, in different areas, the parameters of front feedback check, to check the experience value parameters, using error correcting method, ensure the rationality of the parameter Settings. Experiments prove that the proposed algorithm is compared with the traditional method in the startup delay and parameters such as the load on the server and message overhead and dithering rate, a significant performance improvement.