Abstract: Because of the advantages of efficient , data sharing easily , low requirements for terminal and extremely powerful , cloud computing has occupied an important position in the Internet resources . However , there exists potential safety hazards of the cloud computing restrict the cloud computing development. Nowadays , trust management is one of the hot and difficult spots in the academic research field of the cloud computing security . To solve the dynamism , security and uncertainty of the service resource in the cloud computing environment , the thesis studys and designs a feedback of QoS(Quality-of-Service) trust model based on the service aware theory. In this model , using the time factor of expression to decipt the decay characteristic of trust. To evaluate the trust relationship between users and service providers completely , the model thinks about the direct trust , the recommended trust and the QoS trust. What’s more , since the new joint node in the cloud computing environment , participates the service interaction with a specified probability , hense whitewashing is effectively confine to some extent. By introducing the information entropy and correlation in the evaluation of direct trust and recommended trust , the influence by the malicious review is significantly reduced. At last , simulation results show the trust model is scientific and accuracy.

Key words: cloud security, correlation ratio theory, iInformation entropy, service resource scheduling, Quantity of Service

摘要： 云计算环境下，针对用户无法准确评价服务实体，客观选择服务提供商的问题．结合信息熵和相关比理论来分析实体之间动态信任关系，将服务细粒度化，提出一种基于服务感知的可信QoS（quantity of service）评价模型．在该模型中，引入时间因子刻画信任的衰减特性，综合直接信任、推荐信任以及QoS反馈信任描述实体之间的信任关系；对于新加入节点以某先验初始概率参与服务交互，能在一定程度上有效地抵御漂白攻击；在推荐信任度量与直接信任度量计算中引入信息熵和相关比理论测度，能有效地减弱恶意评价对信任度量的影响．模拟实验表明，该模型在很大程度上满足用户选择服务的需求，并提高资源调度的成功率，有效地保障服务资源调度的安全性．

关键词: 云安全, 相关比理论, 信息熵, 服务资源调度, 服务质量