Ensuring the security of wireless networks using neural networks

In today's world, where wireless technologies are becoming increasingly widespread, ensuring their security is of critical importance. Threats and vulnerabilities in information systems can lead to serious consequences, such as confidential information leaks, financial losses, and disruption of infrastructure systems. Traditional methods, such as signature analysis, have limitations and are not always capable of detecting new attacks or changes in known intrusion patterns. In this context, the use of artificial neural networks (ANNs) to enhance wireless network security is becoming a significant and promising area of research and development. One promising approach is the use of a multilayer perceptron to detect anomalies in network traffic. A multilayer perceptron can learn from historical data and identify complex patterns in traffic that may indicate the presence of network attacks. This study examines modern network attacks, such as DDoS, Man-in-the-Middle, phishing, and protocol vulnerability attacks, identifying characteristics that enable the classification of various attack types. A method is proposed to detect such anomalies using a neural network capable of adapting to new threats and effectively detecting them. Special attention is given to network traffic parameters, such as packet types and lengths, IP address sources and destinations, as well as characteristic signs of anomalies. Methods for detecting deviations have been developed for each parameter, allowing the neural network to make informed decisions.