The relevance of this research is determined by the increasing demands on the performance of multiprocessor computer systems, which are widely used for processing large-scale data and solving complex computational tasks. Uneven load distribution among processors often leads to resource underutilization, overload of certain nodes, and, consequently, a decrease in overall system efficiency. The subject of the study is the process of load balancing in multiprocessor computer systems using metaheuristic optimization methods. The purpose of the work is to develop and analyze a mathematical model of load balancing based on the Particle Swarm Optimization (PSO) method, aimed at improving system performance and resource utilization efficiency. The paper presents a mathematical model of the optimization process for task distribution across processors, considering their performance and current workload. The results of simulation experiments confirm a reduction in the average execution time of computational tasks and an improvement in load uniformity when applying PSO, compared to traditional approaches. The conclusions highlight that the use of PSO is an effective and feasible solution to the load balancing problem in multiprocessor computer systems. The proposed approach can be applied in cloud infrastructures, distributed environments, and high-performance computing systems, where efficient resource allocation is a critical requirement.
Kuchuk et al. (Sun,) studied this question.