基于SPD的网络演化仿真研究

Network Evolution Modeling and Simulation Based on SPD

复杂网络结构演化研究中多探讨“如何形成”,而忽视了“为什么这样形成”的问题。基于合作演化的角度,利用空间囚徒困境理论,对社会网络中的个体进行了分类,并建立网络演化中个体选择的微观动力学机制,建立了社会网络的结构演化模型。使用多主体系统仿真工具Repast进行了仿真。利用度分布、聚集系数、平均最短路径及社会总收益作为演化判据,给出了网络演化的仿真结果。结果表明合作机制下的演化网络展现出明显的小世界特性,说明合作机制可以在一定程度上解释现实网络形成的原因。并且指出对于社会整体来说,即使在合作者较少的情况下,也能够通过社会关系的改善极大的提高社会的总体收益。

The researches on evolution of complex network focused on ＂how to form＂, but ignored ＂why form like this＂. From the view of cooperation evolution, based on spatial game theory, the individuals are classified into two types, and the micro - mechanism of individual choice is analyzed. At last, the model for social network structure evolution is built. The simulation is implemented by the multi - agent system simulation tool Repast. By the criterion of degree distribution, clustering coefficient and average shortest path, the result is showed. The conclusion shows that the evolutionary network has obvious small world property, and the cooperation evolution can explain why the real network forms like this to some extent. Another conclusion is that high total social profit will be gained through the improvement of the network even if the cooperators are few.