系统仿真学报 ›› 2019, Vol. 31 ›› Issue (9): 1790-1801.doi: 10.16182/j.issn1004731x.joss.17-0320

• 仿真建模理论与方法 • 上一篇    下一篇

小世界网络下病毒式信息传播的仿真分析

李锋1, 魏莹2   

  1. 1. 华南理工大学工商管理学院,广东 广州 510640;
    2. 暨南大学企业管理系,广东 广州 510632
  • 收稿日期:2017-07-04 修回日期:2017-08-24 发布日期:2019-12-12
  • 作者简介:李锋(1975-),男,河北邢台,博士,副教授,研究方向为运营管理;魏莹(1977-),女,江西萍乡,博士,副教授,研究方向为运营管理、库存管理。
  • 基金资助:
    国家自然科学基金(71572070),广东省自然科学基金(2014A030313262)

Simulation Analysis of Viral-Style Information Diffusion in Small World Networks

Li Feng1, Wei Ying2   

  1. 1. School of Business Administration, South China University of Technology, Guangzhou 510640, China;
    2. Department of Business Administration, Jinan University, Guangzhou 510632, China
  • Received:2017-07-04 Revised:2017-08-24 Published:2019-12-12

摘要: 鉴于小世界特征的复杂网络的普适性,以经典的病毒式信息传播SIR模型为对象,采用多智能体建模与仿真方法分析小世界网络的结构属性与信息扩散的网络覆盖率指标之间的关系通过仿真,计算了小世界网络Watts-Strogatz算法中不同参数设置对信息扩散的影响。计算结果不仅得到了数学分析无法得到的网络覆盖率随机分布的双峰形态,而且计算得到了其与网络生成算法参数之间的关系。确定了信息扩散的网络覆盖率指标与小世界网络的平均路径长度、网络直径、平均聚类系数等网络结构指标之间的强相关性和函数表达式。

关键词: SIR模型, 信息扩散, 网络覆盖率指标, 小世界网络, 多智能体建模仿真

Abstract: Since the small world network is one of the most common complex networks in the real world, this paper focuses on information diffusion on the small world network platform. In order to explore the relationship between structural features of the small world network with a well-accepted feature for information diffusion, network coverage ratio named, the methodology of multi-agent based modeling and simulation is introduced, instead of mathematical modeling. Through simulation, the impact of only two parameters in the small world network generation algorithm, that is Watts-Strogatz algorithm, is analyzed, where the process of information diffusion acts as the SIR model. The simulation results get not only the bimodal pattern of distribution of the network coverage, but also its relationship with characteristics of the network generation algorithm. The regression function of network coverage ratio variable with the complex network characteristics such as average path length, network diameter, and average clustering coefficient is built.

Key words: SIR model, information diffusion, network coverage index, small world network, multi-agent modeling and simulation

中图分类号: