Adaptive Cloud Load Balancing Using Opposition- Learning- Artificial Bee Colony Optimization

Abstract

The increasing complexity and scale of cloud computing environments demand intelligent strategies for managing workloads across distributed resources. Load balancing plays a pivotal role in ensuring that computational tasks are efficiently allocated to available servers, thereby improving performance, reducing latency, and maintaining service reliability. This paper introduces an enhanced metaheuristic approach for cloud load balancing by combining the Artificial Bee Colony (ABC) algorithm with Opposition Learning (OL). While ABC provides a biologically inspired mechanism for optimization, its performance may degrade in high-dimensional or rapidly changing scenarios due to premature convergence. The integration of OL enhances the exploration capability of the algorithm by considering opposite candidate solutions during the search process, increasing the chances of escaping local optima. This research presents a comprehensive overview of the OLABCA-LB framework, proposed as an effective solution for dynamic load balancing within cloud environments. The study introduces a detailed mathematical model along with the key parameters integrated into the fitness function design, which collectively guide the balanced allocation of computational tasks across physical hosts and virtual machines. The paper also outlines the simulation environment, describing the setup and configuration used to evaluate the proposed algorithm. Performance analysis is conducted under varying conditions, including different task volumes, instruction lengths, and scales of virtual machine deployment, to validate the robustness and scalability of the OLABCA-LB approach