Wireless Intrusion Detection Using Shallow Neural Network Models

Abstract

Wireless systems, due to their nature lack the security features against various attacks such as jamming and eavesdropping. In order to successfully detect such attacks that occur in a wireless network, Artificial Intelligence (AI) models which continuously monitor wireless statistic records are used. On this context, this paper proposes an artificial neural networks based Wireless Intrusion Detection System (WIDS) for 802.11 (Wi-Fi) wireless networks, where the model is trained with public Aegean Wi-Fi Intrusion Dataset (AWID-2).