DSRTrust: A Dynamic Trust Model of Distinguishing Service and Recommendation for Internet-based Virtual Computing Environment

Abstract

Internet-based virtual computing environments (iVCE) are open, anonymous and dynamic in nature. Such characteristics bring about threats and vulnerabilities in providing trusted services and improving resource utilization. Therefore, a dynamic trust model for distinguishing service and recommendation is proposed. In this paper, we analyze multidimensional decision factors related to the evaluation of autonomous node, such as user satisfaction, reward function, punishment function and time decay function. According to the network connection degree of node, our model assigns a new trust weight that specifically describes the relationship between network and trust in iVCE. We then propose a dynamic quantitative model for measuring different kinds of trust. Simulation results indicate that our model can effectively cope with malicious behavior and exhibits evident advantages in resource utilization compared with existing models.