A New Multi-Objective Optimization Algorithm to Solve the Load Balancing Problem in Mobile Cloud Computing

Abstract

Mobile Cloud Computing (MCC) has emerged as a promising paradigm to overcome the computational and energy limitations of mobile devices by offloading intensive tasks to the cloud. However, determining optimal task offloading and scheduling strategies remains a challenging multi-objective optimization problem due to the heterogeneous nature of cloud resources and constraints such as execution time, energy consumption, and bandwidth. This paper proposes a novel Multi-Parallel Objective Imperialist Competitive Algorithm (MPICA) to efficiently address task scheduling in MCC environments. By leveraging parallel processing, MPICA enhances exploration and exploitation in the solution space, leading to improved convergence speed and load balancing. The performance of MPICA was evaluated against three benchmark algorithms: Round Robin (RR), Genetic Algorithm (GA), and the standard Imperialist Competitive Algorithm (ICA). Simulation results demonstrate that MPICA achieves up to 25% reduction in makespan and 18% improvement in energy efficiency, while maintaining better scalability in large-scale task sets. These findings highlight the potential of MPICA as a robust and scalable solution for multi-objective task scheduling in MCC scenarios.