ANALYSIS OF GENETIC ALGORITHM APPLICATION ASPECTS IN MODERN NETWORKS

Abstract

Due to the increasing volume of data and the growing complexity of network tasks, there is a need for methods capable of finding optimal solutions under dynamically changing conditions. Over the past decade, there has been a rapid development of methods based on evolutionary principles, which highlights the relevance of a thematic review of studies focused on the application of the Genetic Algorithm (GA) in communication networks. This review aims to systematize and generalize information, as well as identify development trends and recent advancements in this field. The study examines thematic connections and the evolution of GA research by analyzing keywords and their interrelations in scientific publications. Three primary GA operators are reviewed, along with the probability values of crossover and mutation, which directly impact the final solution in various studies related to modern network operations. Based on the conducted analysis, it is noted that the choice of operators and parameters in GA has a fundamental impact on their functionality and the efficiency of problem-solving. It is crucial to consider that optimal parameters may vary significantly depending on the task and data type. Effective configuration requires thorough experimentation and adaptation of parameters to specific conditions, making the direct application of parameters from other studies not always advisable. A classification of GA has been developed based on the following criteria: parameter adjustment principles for crossover and mutation, type of fitness function, algorithm type, and the nature of network tasks. The proposed classification helps to elucidate the fundamental principles of the algorithm's operation, enhancing its practical application efficiency and contributing to the systematization of relevant knowledge.