Retracted: Multilayer Statistical Intrusion Detection Model for Wireless Network

The purpose of an intrusion detection system (IDS) is to monitor network traffic for any unusual behaviour. IDSs have been studied using a variety of methodologies, but due to the volume of data, such approaches have not been able to produce results with high precision. The disappointing outcomes of conventional intrusion detection systems have led to the statistical approach that is presented in this study. Our proposed method uses two consecutive layers of security; the very first layer is at the edge of the networks, close to the devices. It employed the Fuzzy Vikor classifier, which divides traffic into legitimate and malicious traffic, in order to streamline and expedite the decision-making process. While the second layer is in the cloud, it uses a Multi-Class Q-Learning classifier to primarily handle suspicious traffic. The implementation results show that, in addition to achieving a high Precision rate, our proposed multi-layer protection model also improved the values of TPR, TNR, FPR, and FNR. These values were 99.6%, 99.4%, 99.5%, 98.5% and 99.6% percent for the Normal, Flooding, Scheduling, Grayhole, and Blackhole attacks, in both.