We consider the problem of detecting the insiderbased attacks in the form of jammers in a single-hop wireless LAN environment, where jammers have the inside knowledge of frequency hopping patterns and any protocols used in the wireless network. We propose a novel jammer model in which the jammers are modeled by the number of channels that they can jam simultaneously. We further propose the novel concept of an atomic jammer which is the basic component necessary to deal with stronger jammers. To deal with atomic jammers, we propose a class of novel protocols called alibi protocols. The basic idea of the alibi protocols is to exploit one major limitation of the atomic jammers: they cannot jam two channels at the same time. Therefore, honest nodes in the network can occasionally switch to another channel, called the alibi channel, to transmit proofs for their honesty witnessed by some other honest nodes. We specify a necessary condition and desired properties such as detection time, false alarms and miss detections of this class of protocols. We prove that with high probability the detection time of these protocols is O(n ln(n)). We also propose some more practical alibi-based protocols such as 1-propagation and 1-gossiping and prove their desired properties. We further extend our work to the lossy channel model. The simulation results in ns2 confirm our analysis. The overall results of these protocols show a promising research direction to deal with insider-based jamming attacks.
展开▼
