多Agent系统安全信任问题研究

Research on the Security and Trust based on the Mult-Agent

随着网络的普及和计算机性能的提高,智能Agent技术以其高度的灵活性和智能型等特征在解决未来复杂问题方面具有独特的优势。本文以多Agent合作为场景提出了基于关系网的信任模型。提出了基于信任模型的分层信任推理算法,实现了一个能对信任进行推理的框架,并且给出了信任评估模型以及度量信任的机制。

Along with the popularization of the network and the upgrade of computer performance, the diversification and complication of the information system make people look for more the intelligent and autonomous approach to resolve the complicated problem. The relation net model with acquaintance’s mechanism is developed to describe he trust among Agents. The layered reasoning algorithm is used to calculate the rust degree, and trust valuation and measurement mechanism are provided. On his foundation, a reasoning framework of trust is implemented