An improved Grey Wolf Optimizer (iGWO) for Load Balancing in Cloud Computing Environment

Load balancing in any system aims to optimize throughput, resource use, imbalance load, response time, overutilization of resources, etc. An efficient load balancing framework in cloud computing environment with such features may improve overall system performance, resource availability and fulfillment of SLAs. Nature-inspired metaheuristic algorithms are getting more popularity day by day due to their simplicity, flexibility and ease implementation. The success and challenges of these algorithms are based on their specific control parameter selection and tuning. A relatively new algorithm motivated by the social hierarchy and hunting behavior of grey wolves is Grey Wolf Optimizer (GWO), which is having least dependency on the control parameters. In the basic GWO, 50% of the iterations are reserved for exploration and others for exploitation. The perfect balance between exploration and exploitation is overlooked in GWO. The impact of perfect balance between two guarantees a near optimal solution. To get over this problem, an improved GWO (iGWO) is proposed in this paper, which focuses on the required meaningful balance between exploration and exploitation that leads to an optimal performance of the algorithm. Simulation results based on exploitation and exploration benchmark functions and the problem of load balancing in cloud demonstrate the effectiveness, efficiency, and stability of iGWO compared with the classical GWO, HS, ABC and PSO algorithms.