Channel Detecting Jamming Attacks on Rendezvous Algorithms for Cognitive Radio Networks.

摘要

Traditional spread spectrum techniques such as Frequency Hopping Spread Spectrum (FHSS) and Direct Sequence Spread Spectrum (DSSS) have been used for anti-jamming solutions in wireless broadcast communication. However, these systems generally have the fundamental limitation of a prior key sharing between a sender and receiver. If a jammer is a compromised receiver, then it uses the secret key to jam the entire wireless communication. To address this problem, new enhancements allow a sender and receiver to independently generate random channel hopping (CH) or frequency hopping (FH) sequences so that it is unfeasible for a jammer to compute the same sequences. These schemes can provide fast rendezvous or key exchange methods for Cognitive Radio Networks (CRNs) called blind rendezvous algorithms in which a sender and receiver have no prior knowledge of a shared key, time synchronization information, or common control channels (CCCs).;However, we present new channel detecting jamming attacks (CDJAs) against these enhancements for CRNs. For most rendezvous algorithms, our channel detecting jammer can compute the same sequences as the sender's by utilizing the properties of blind rendezvous schemes. We investigated the state-of-the-art blind rendezvous algorithms for CRNs to demonstrate the effectiveness of our CDJAs. Through simulations, we show that CDJAs can significantly reduce their rendezvous probability for both the symmetric and asymmetric rendezvous systems. Thus, our CDJAs are a major security problem for most blind rendezvous algorithms since any secondary user or even group of users in CRNs can easily be denied access to the network with high probability. To mitigate this problem, we revisit the Random rendezvous scheme to increase the rendezvous probability against CDJAs. Overall, the Random scheme vastly outperforms these representative blind rendezvous algorithms for both the symmetric and asymmetric models when there are security concerns about a channel detecting jammer. A new partially random algorithm is shown to outperform all others for the asymmetric system.