现有的网构软件系统信任模型缺乏对软件实体信任协作过程的分析,以及对信任协作结果的有效预测。为此,运用贝叶斯博弈对软件实体的信任协作关系进行研究。阐述网构软件系统中存在的安全问题,给出网构软件的演化模型,在此基础上,采用贝叶斯博弈理论对系统中软件实体的信任协作过程进行建模。针对博弈过程中信息不完整的问题,通过引入自然选择的概率,利用Harsanyi转换得到更符合实际情况的贝叶斯纳什均衡解,对信任协作结果进行有效预测。实验结果表明,信任协作结果与软件实体与自然选择概率的判断有关,通过调整博弈模型中的相关参数可以促进软件实体有效地进行协作。
In order to solve the problem that the existing trust models for intemetware lack the analysis and prediction for the trust cooperation process of the software entities, this paper researches the trust collaboration relation of software entities by using Bayesian game. It points out the security problem of internetware, puts forward an evolution model of internetware, and models for the trust cooperation process of the software entities by Bayesian game. It uses Harsanyi conversion to get the Bayesian Nash Equilibrium(BNE), and predicts for the result of the cooperation game. In order to promote the cooperation of the software entities, the relevant parameters in the game model is analyzed and discussed through the experiments. Result shows that it can improve the overall security of the internetware system.