在云计算中,终端用户对云资源软硬件的影响和破坏远比目前用户利用因特网进行资源共享要严重的多,因此相对于用户的身份,用户的行为是否真实可信是云计算研究的一个重要内容。提出了户行为认征的机制,包括行为认证集的确立、行为证据的获得、行为认证的策略,建立相应的随机Petri网模型,并通过量化瞬时变迁的实施概率来模拟不同的认证选择,利用系统达到稳定状态时位置中的平均标记数来评价认证效果,为云计算的安全应用奠定理论和实践基础。
The impact to the hardware/software resource generated by end users in cloud computing is much more significant than in the Internet resource sharing scenario. Therefore the authentication of user behavior becomes a key aspect in cloud computing research community. A scheme for user behavior authentication was purposed, including the establishment of authentication set, the acquirement of user behavior evidence, and a scheme for behavior authentication. A Stochastic Petri Net Model of the system was built to analyze the system performance by computing the steady state marking distribution.