Hybrid task scheduling strategy for cloud computing by modified particle swarm optimization and fuzzy theory

As the world is progressing towards more efficient computing and faster approaches, cloud computing is a popular computing model to such increasing requirements. In order to provide cost-effective executions in cloud environment, appropriate task scheduling strategy is necessary. This paper proposes a hybrid task scheduling algorithm named FMPSO that is based on Fuzzy system and Modified Particle Swarm Optimization technique to enhance load balancing and cloud throughput. FMPSO strategy at first considers four modified velocity updating methods and roulette wheel selection technique to enhance the global search capability. Then, it uses crossover and mutation operators to overcome some drawbacks of PSO such as local optima. Finally, this schema applies fuzzy inference system for fitness calculations. The input parameters for the proposed fuzzy system are length of tasks, speed of CPU, size of RAM, and total execution time. By adding these fuzzy systems, FMPSO strategy achieves the goal of minimizing the execution time and resource usage. We evaluate FMPSO algorithm using the CloudSim toolkit and simulation results demonstrate that the proposed strategy has a better performance in terms of makespan, improvement ratio, imbalance degree, efficiency, and total execution time comparing to other approaches.