复杂网络上重复囚徒困境博弈的研究.pdfVIP

Vo1. 26 No.7 第 26 卷第 7 期 计算机应用研究 2{朋年7 月 Application Research of Computers Jul. 2仪渺 复杂网络上重复囚徒困境博弈的研究* 王伊蕾\杨洪勇a ，李涛b ( -t-*大学 a. 计算机科学与技术学院; b. 网络中心，山东烟台却崎25) 摘要:为了研究博弈理论在复杂网络上的影响，对Zachary 网络应用重复囚徒因境博弈的思想提出了两种网 络演化算法，即随机算法和伪度优先算法，并对才发发博弈后的复杂网络特性进行了分析，其中包括皮分布特性和 聚集系纸。最后采用 Newman 算法，对经过伪1t优先算法后的网络进行社团结构分析，发现伪度优先算法可以 明显改善网络的社团结构。结果表明级过两种算法后的网络，底分布依然服从来律分布，聚集系数均有所增加， 而且伪皮优先算法增加幅度很大。 关键词:囚徒困境;纳什均衡;一报还，时偏报策略:复杂网络 中阁分提号: N94; 四24 文献标志码:A 文章编号: 1ω1-3695(2:α9 )07-2683-03 doi: lO. 3969/j. issn. 1∞1δ695. 棚9.07.079 Research of iterated prisoner s dilemma game on complex network WANGYi刷lei. ， YANG Hong翩yonga ， U Taoh (α. Schω，lofCo呻脚r Sc旬næ T.础阳logy， b. Network C阴阳， Ú础ng Univeritr , Yantαi Sharulong 剧。25 ，刷刷) Abstract: To 阳dy the impact of game theory on complex network ,this paper put forward two network evolutionary algorithms- random algorithm and p脚udo value first algorithm that apply iterated prisoner s dilemma gωne theory on Zachary network. And studied the network characters after iterated games , which included de伊哩e distribution and its cluster coefficient characters. At last adopted Newman algorithm to analysis network community structures after pseudo value first algorithm , found that this al- gont阳n could improve network community structures. The results show that degree character of networks after the two a!go- rithms also obey sca!e-free distribution and their cluster coefficient inc陀ase with games go on while cluster coefficient increase a lot after pseudo value first algorithm. Key words: pnsoner s