IMPLEMENTATION OF AN INTELLIGENT SYSTEM FOR EFFECTIVE TRAFFIC DISTRIBUTION IN COMPUTER NETWORKS

Abstract

The solution to the problem of efficient distribution of traffic in computer networks using Artificial Intelligence algorithms is presented. The problem of effective distribution of traffic in computer networks is considered one of the key problems in the field of network interaction organization. It is very pivotal in the field of traffic management to increase the efficiency of using the existing channels and the quality of services provided to end users. Currently, this problem has become more relevant, as the scope of its application has expanded, which now includes both policy enforcement systems and the field of information security. The paper considers different Artificial Intelligence algorithms that can be used for the efficient distribution of network traffic: Support Vector Machine (SVM), Random Forest (RF) and Artificial Neural Networks (ANN), as well as their advantages and disadvantages are presented. After features used to classify network traffic were analyzed, the best set of network traffic features was selected. Step-by-step testing of an intelligent system for effective network traffic distribution was performed according to the testing scheme. A comparative analysis of the recorded results of the mentioned Artificial Intelligence algorithms and a reasonable choice is made. Based on the chosen Random Forest (RF) algorithm, an intelligent system for the effective distribution of network traffic was designed.