JaDe: Low Power Jamming Detection Using Machine Learning in Vehicular Networks

Vehicular networks, like any other wireless networks, are prone to jamming due to the inherent nature of the wireless environment. In this letter, we consider an attacker who jams the network whenever he observes that a packet is being transmitted, i.e., some communication happening in the network. Such a strategy leads to an increase in the packet drop rate and latency of the network. This causes serious disruption of delay-sensitive applications that can lead to life endangering situations and therefore such an attack needs to be addressed. In this letter, we propose a mechanism that uses a support vector machine to detect the presence of a jammer in the network. We obtain jointly sufficient statistics of packet drop probabilities and use them to generate the training data. The results demonstrate the effectiveness of the proposed detection system.