A Greedy Stochastic Diffusion Search Based Fuzzy Scheduling in Cloud

Abstract

Users in cloud environment pay per use for the resources utilized. Security issues direly affect the cloud services due to which the performance of the cloud systems also gets affected. Hence, for robust security in data and systems in cloud, the onus is largely on service providers. The most important component of cloud computing is job scheduling. For distributing precedence to subtasks that achieve varied makespan in heterogeneous computing system based on the different procedures adopted by different job scheduling algorithms. Additionally, every resource assigned to a task may consume variable amounts of energy. The problems of solving the cloud scheduling problem are Non-deterministic Polynomial (NP) hard problem. The minimization of makespan is the target of most investigations. This work considers both makespan as well as energy consumption. For optimizing these tasks, the greedy and the Stochastic Diffusion Search (SDS) algorithm are combined in job scheduling in this work. Fuzzy reasoning is combined with greedy SDS mechanism by the proposed methodology for optimizing scheduling. The greedy SDS method is used for exploiting the fuzzy solution space and combining the benefits of both the heuristics and evading the drawbacks is the basic tenet of the proposed approach. It has been shown that in terms of both the consumption of energy and makespan, the proposed algorithm performs better than the existing algorithms.