JoWUA

Abstract

Wireless networks are an important aspect of communication technologies that avoid the cost and burden of cable installation. They play a vital role in our everyday lives. However, these wireless networks have some limitations which can be exploited by malicious users to capture transmitted information or cause disruptions in communications. A Radio Frequency Jamming (RF-Jamming) attack is one such type of attack that interferes with authentic wireless signals by reducing the signalto- noise ratio. These types of attacks pose serious threats to many applications especially the safetycritical ones such as self-driving cars. Hence, it is crucial to institute countermeasures to prevent these attacks and establish a reliable communication system. Furthermore, to take the appropriate steps for the protection against such attacks, it is important to know the type of jamming attack that a network has been exposed to. In other words, in addition to detection, the classification of these attacks is also necessary. Therefore, in this paper, we tackle this problem and propose a machine learning-based classification technique for different types of jamming attacks. We simulate the jamming scenario in wireless ad-hoc networks using the network simulator ns-3 and use the data collected from the simulation to train and evaluate different algorithms. We compare the accuracy of each algorithm and provide the results that showcase that the classification of jamming attacks can be done with very high accuracy using the Gradient Boosting Algorithm.