Modified Moth-Flame Optimization Algorithm for Service Composition in Cloud Computing Environments

Cloud computing service composition integrates services, distributed and diverse by nature, into an integrated entity that can meet a user's requirement with better effectiveness. However, some obstacles regarding high latency and suboptimal Quality of Service (QoS) still exist in a dynamic multi-cloud environment. This study addresses the limitations of traditional optimization algorithms in service composition, specifically the premature convergence and lack of population diversity in the Moth-Flame Optimization (MFO) algorithm. We propose the modified MFO algorithm with a new mechanism called Stagnation Finding and Replacement (SFR) to enhance the diversity of the population. It finds the static solutions based on a distance metric from globally optimal representative solutions and replaces them. MFO-SFR drastically improved all QoS metrics, such as response time, delay, and service stability. Empirical evaluations prove that MFO-SFR outperforms the baseline methods of multi-cloud service composition. It provides acomputationally efficient and adaptive solution to cloud service composition problems, ensuring better resource utilization and higher user satisfaction in dynamic multi-cloud environments.