Design of Intrusion Detection System for Wireless Ad Hoc Network in the Detection of Man In The Middle Attack using Principal Component Analysis Classifier Method Comparing with ANN classifier

This work aims to construct a man-in-the-middle attack detection system for wireless adhoc networks using a classifier based on principal component analysis. Its effectiveness will be contrasted with that of an ANN-based intrusion detection system. The network dataset design of an IDS was simulated and produced with a an-in-the-middle attack using the NS2 tool. The 12 network layer features were supported by the NS2 simulation environment. In the machine learning platform, this IDS's model and testing were done. SPSS is used to analyse the two IDSs. For each group 19 samples are taken. The performance of the IDSs is demonstrated by significance p0.005, mean accuracy, and detection rate. According to the experimental findings, the IDS, which was created using principal component analysis, has an accuracy rate of 76 percent and a detection rate of 74 percent. The accuracy and detection rate of the IDS, which used the ANN algorithm, are both 100%. The IDS developed using the ANN Algorithm appears to perform significantly better than the IDS that uses a principal component analysis, it is found.