Trust Model Considering Uncertainty and Dynamics in Cloud Computing

With the rapid popularization of cloud computing, the increase in cloud traffic and resources has come with increased uncertainty. Therefore, the consideration of uncertainty and dynamics is becoming more important for the effective management of cloud resources. This article aims to establish a more accurate and reliable service evaluation model suited for a dynamic and changeable network environment, which provides users with more precise service selections and recommendations. Cloud service (CS) trust evaluation is conducted from subjective and objective perspectives in this paper. First, a set of objective criteria evaluation cloud is generated based on multivalued monitoring data, which can more accurately reflect the dynamic performance of CSs. Second, similar recommenders are identified based on the similarity of the objective criteria evaluation cloud of the current user and recommenders. Then an objective assessment is employed as a benchmark to remove malicious or potentially biased subjective recommendations by comparisons with corresponding subjective assessment characteristics. Finally, the objective trust evaluation cloud is converted into a trust degree expressed by the probabilistic linguistic term set (PLTS) method. Additionally, the subjective trust degree and trust preference defined by the trustor are combined to evaluate the trustworthiness of the CSs. Using the PLTS method to scale the trust of each criterion of CSs can systematically describe the dynamic performance of CSs, and consider the uncertainty and vagueness of trust. The experimental results demonstrate its effectiveness, feasibility and better accuracy compared with other methods.