Detecting Flooding, Impersonation and Injection Attacks on AWID Dataset using ML based Methods
Mayank Agarwal · 2022
The widespread use of Wi-Fi networks has also made them a preferred target to a wide range of security assaults. Assailants are employing more advanced techniques to launch attacks, which is changing the nature of the attacks. Using the publicly accessible Aegean Wi-Fi Intrusion Dataset, the authors describe a machine learning (ML) based Wireless Intrusion Detection System (WIDS) for identifying flooding, impersonation, and injection attacks in Wi-Fi networks (AWID). The benefit of ML-based IDS is that they can decipher complicated patterns from the data, allowing them to distinguish between patterns of legitimate traffic and malicious traffic. On the AWID dataset, the authors contrast and compare the results of Logistic Regression (LR), AdaBoost, Naive Bayes (NB), Long Short-Term Memory (LSTM), Decision Tree (DT), and Random Forest (RF). The authors have used data preparation techniques on the AWID dataset’s null values. The trials showed that RF and DT outperformed other ML approaches for the detection of flooding, impersonation, and injection attacks in terms of accuracy, precision, recall, and F-measure. Our proposed methods outperform the others by a wide margin, as demonstrated by a comparison with contemporary methods that have been employed in the literature.