Optimized Load Distribution in Cloud Computing Using Random Opposition-Based Coati Optimization Algorithm (RO-COA)

Abstract

Cloud computing has become an integral part of modern IT infrastructure, providing scalable, on-demand resources to users worldwide. Efficient task scheduling and load balancing are critical challenges in cloud environments, as uneven distribution of workloads can lead to underutilized virtual machines, overloaded servers, increased energy consumption, and reduced Quality of Service. Traditional heuristic and metaheuristic optimization methods often struggle to find global optima due to premature convergence or limited exploration capabilities. This paper introduces the Random Opposition-Based Coati Optimization Algorithm (RO-COA), a bio-inspired optimization technique designed specifically for multi-objective cloud resource scheduling. The algorithm leverages the cooperative foraging behavior of coatis and incorporates a random opposition-based learning mechanism to improve both exploration and exploitation during the search process. RO-COA evaluates each candidate solution alongside its dynamically generated opposite, ensuring diversity in the search space and preventing the algorithm from being trapped in local minima. Experimental simulations demonstrate that RO-COA effectively balances CPU and memory utilization, minimizes task makespan, reduces energy consumption, and improves overall system performance compared to conventional optimization techniques. The approach offers an adaptive, robust, and scalable solution for cloud resource management, making it suitable for both homogeneous and heterogeneous cloud infrastructures.