Modified Artificial Bee Colony Algorithm for Load Balancing in Cloud Computing Environments
Qian Li, Xue Wang · International Journal of Advanced Computer Science and Applications · 2024
Task scheduling in cloud computing is a complex optimization problem influenced by the ever-changing user requirements and the different architectures of cloud systems. Efficiently distributing workloads across Virtual Machines (VMs) is critical to mitigate the negative consequences of inadequate and excessive workloads, such as higher power consumption and possible machine malfunctions. This paper presents a novel method for dynamic load balancing using a Modified Artificial Bee Colony (MABC) algorithm. The ABC algorithm has exceptional competence in solving complex nonlinear optimization problems based on bee colonies' foraging behavior. Nevertheless, the traditional version of the ABC algorithm cannot effectively use resources, resulting in a rapid decline in population diversity and an ineffective spread of knowledge about the best solution between generations. To address these limitations, this study integrates a genetic model into the algorithm, enhancing population diversity through crossover and mutation operators. The developed algorithm is compared with the prevailing algorithms to confirm its effectiveness. The results of the proposed MABC algorithm for the load balancing method are compared with the current ones, and it is observed that this algorithm is more beneficial in terms of cost and energy as well as resource utilization.