A probabilistic framework for jammer identification in MANETs

摘要

Cyber maneuver is envisioned as a way of increasing the robustness of imperfect systems by creating and deploying mechanisms that continually change a system's attack surface. A particularly useful strategy against jamming attacks in Mobile Ad Hoc Networks (MANETs) entails using cyber-maneuver keys to supplement higher-level cryptographic keys. Such keys can be periodically changed either in a proactive fashion - to defeat cryptanalytic efforts by external attackers - or in a reactive fashion - to exclude compromised internal nodes. In order to enable effective reactive rekeying, it is critical to correctly identify compromised nodes. In this paper, we propose a probabilistic framework for identifying jammers, based on the location of both jammed and non-jammed nodes. We are interested in finding the smallest set of nodes that need to be excluded to stop the attacks in a multi-jammer scenario. We show that this problem is NP-hard, and propose a polynomial-time heuristic algorithm to find approximate solutions. Experiments show that our approach works well in practice, and that the algorithm is efficient and achieves good precision and recall.