DETECTION AND CLASSIFICATION OF FLOODING ATTACKS IN WIRELESS ADHOC NETWORKS USING MACHINE LEARNING

Abstract

In developing wireless adhoc networks, several sorts of attacks represent serious security problems.Various forms of attacks are currently being carried out against various services and resources, with the goal of compromising their availability, confi dentiality, and integrity. Flooding based Denial of Service attacks has a serious impact on Wireless Local Area Networks. It may result in clients being denied service. This is a severe problem since it compromises one of the services offered by cyber security such as availability. In ad-hoc networks, it is crucial to detect and block such attacks in a timely manner. The objective of this thesis is to accurately detect and categorise flooding attacks such as TCP, UDP, or ICMP. This thesis additionally categorises ad ditional assaults, such as U2R, R2L, and probe .The suggested method utilises various supervised machine learning algorithms, including SVM, KNN, Naive Baye’s, Deci sion Tree, and Random Forest. Classification is performed using the NSL KDD data set. The outcome demonstrated that RF classifier provides the highest performance accuracy