一种基于局域世界的可调簇系数网络演化模型

Local-world-based evolving network model with tunable clustering coefficient

网络拓扑的结构是网络协议和网络算法设计的基石,合理的网络拓扑模型无疑可以指导协议的设计及优化。研究了现实网络的拓扑特性,引入局域世界的概念,建立了一种基于局域世界的可调簇系数网络演化模型(LWTC)。该模型全面考虑了局域内以及局域外部加点、偏好链接以及链接随机消亡等诸多因素,并引入三角形形成机制调节模型的簇系数。通过MATLAB和C++对模型进行仿真分析发现,该模型动态模拟了网络的演化过程,其度分布不仅与理论推导值符合,簇系数也优于局域世界演化模型,并且通过合理调节参数可以改变模型的簇系数,较好地刻画了现实网络的统计特性。

The structure of topology model is the foundation stone of network protocols, so a reasonable topology can guide the design and optimization of protocols. This paper researched on real network, and put forward a novel model named LWTC based on the local-world. This model took general consideration of adding new nodes, preferential attachment and random death of nodes in or between local-worlds, and introduced triad formation to tune the clustering. The MATLAB and C ＋＋ simulation indicates that not only the degree distribution consistent with the theoretical results, but also the clustering coefficient had a good performance than the local-world evolving network, and the clustering coefficient could be changed by reasonable adjusting of the parameter, so the LWTC model reproduced the properties of real network.