Towards Safer Wi-Fi Networks: Leveraging Neural Networks for Intrusion Detection

The heavy reliance on Wi-Fi networks has made them susceptible to various threats, especially given that the security protocols employed have some flaws that were exploited by malicious entities. In this paper, we introduce the first constituting block of our comprehensive Wi-Fi Intrusion Detection and Resource Management System. The primary objective of this system is to filter and classify incoming threats, thereby enabling subsequent decisive actions within the framework’s subsequent modules. Using the AWID dataset and after selecting subset of attacks a neural network has been trained and optimized with 99.5% accuracy, 95.2% precision, 94.0% recall and 94.5% F1 Score.