DESIGN AND DEVELOPMENT OF A SIMULATION MODEL FOR BANK NETWORK TRAFFIC FILTERING SYSTEMS BASED ON DPI TECHNOLOGY USING THE ANYLOGIC SOFTWARE

Abstract

This article analyzes the characteristics of artificial neural network algorithms and neural network models. Single-layer neural networks are also considered. Although a single neuron is incapable of performing even simple perception processes, the computational power of neural networks emerges when multiple neurons are interconnected within a network. The processes of handling incoming network load using Hamming and Hopfield models are examined. The results of both simulation models demonstrate that network traffic filtering is essential, firstly, for resource optimization, secondly, for protecting service systems from threats and attacks, and thirdly, for reducing request processing time. As a result, both the speed and quality of service are significantly improved.