JADE: Jamming-averse routing on cognitive radio mesh networks

The spectrum sensing capability of cognitive radio (CR) enables a lot of opportunities to wireless networks, but also enables intelligent attacks by malicious players. One attack in this category is reactive jamming, in which the attacker senses the wireless spectrum, decodes parts of packets, and selectively interferes with packets. In so doing, an attacker can reduce energy expenditure and increase stealth while maintaining a high impact. Of the approaches to mitigate jamming, in this work, we focus on the jamming resilient routing in CR mesh networks. To do this we use signal-to-noise-interference ratio (SINR) which reflects the jamming impact. This metric is difficult to measure with commodity radio chipsets that cannot differentiate jamming interference from the received signal. Detecting SINR becomes even harder if reactive jamming is used by an attacker. In this study, we develop a mechanism to estimate SINR under reactive jamming. The estimated SINR information of each wireless link is then used to determine the jamming-averse directivity (JAD) of packets, which improves the routing performance of the victim network. We validate the proposed mechanism with a simulation study, showing that the proposed JAD escorted (JADE) routing dramatically improves routing path discovery performance including path discovery probability, path length, elapsed time for path discovery, retransmission attempts, and path quality under reactive jamming. Among the 200 route requests at 10 different configurations in our simulation, the reactive jammer disrupts the 77.5% of total requests. However, our JADE routing decreases the route discovery failure rate to 7.5% by saving the 96.7% of failed requests.