Proposing a New Algorithm for Predicting Short-Term and Long-Term Trust-ability in Cloud Computing

Despite the huge use of cloud computing, due to its large dimensions and availability for all users, this type of network is weak and vulnerable to malicious attacks. Therefore, as a useful complement to existing security methods, trust management plays a crucial role in discovering suspicious behaviors in the cloud computing network. The critical question is, how can we find ideally and effectively users with suspicious behaviors in these complex environments. In this paper, the Markov chain model has been used to calculate the short-term reliability of users in the cloud network, and the trust management system has been proposed to mitigate the effects of complex environments to calculate the user’s status. Furthermore, a new computational model has been introduced with relevant, practical factors for calculating the long-term trust that reduces the effect of changing environmental parameters in the calculations. In the Markov chain, in each time unit the transition occurs from one state to another, the number of these states can be counted. In this paper, two modes of normal (faultless) and having a risk (damaged, and having a fault) are considered for users. The simulation results show that the proposed algorithm, Markov chain trust management can more effectively detect suspicious behaviors of users in the cloud computing network, and in a meaningful way, provide a better rate of delivery of packets compared to their counterparts, and ultimately provide better security in the cloud computing network. To assess the effectiveness of the introduced Markov chain model, we compared it with two TBID and RFSN models. We have used MATLAB software to compare the performance of the cloud network.