NEIGHBOR SELECTION AND HITTING PROBABILITY IN SMALL-WORLD GRAPHS

• 成果类型: Article
• 署名作者: Sandberg, Oskar
• 署名单位: Chalmers University of Technology; University of Gothenburg
• 刊物名称: ANNALS OF APPLIED PROBABILITY
• ISSN/ISSBN: 1050-5164
• DOI: 10.1214/07-AAP499
• 发表日期: 2008
• 页码: 1771-1793
• 关键词: NAVIGATION; networks

摘要

Small-world graphs, which combine randomized and structured elements, are seen as prevalent in nature. Jon Kleinberg showed that in some graphs of this type it is possible to route, or navigate, between vertices in few steps even with very little knowledge of the graph itself. In an attempt to understand how such graphs arise we introduce a different criterion for graphs to be navigable in this sense, relating the neighbor selection of a vertex to the hitting probability of routed walks. In several models starting from both discrete and continuous settings, this can be shown to lead to graphs with the desired properties. It also leads directly to an evolutionary model for the creation of similar graphs by the stepwise rewiring of the edges, and we conjecture, supported by simulations, that these too are navigable.