Thwarting Control-Channel Jamming Attacks from Inside Jammers

Coordination of network functions in wireless networks requires frequent exchange of control messages among participating nodes. Typically, such messages are transmitted over a universally known communication channel referred to as the control channel. Due to its critical role, this channel can become a prime target of Denial-of-Service (DoS) attacks. In this paper, we address the problem of preventing control-channel DoS attacks manifested in the form of jamming. We consider a sophisticated adversary who has knowledge of the protocol specifics and of the cryptographic quantities used to secure network operations. This type of adversary cannot be prevented by antijamming techniques that rely on shared secrets, such as spread spectrum. We propose new security metrics to quantify the ability of the adversary to deny access to the control channel, and introduce a randomized distributed scheme that allows nodes to establish and maintain the control channel in the presence of the jammer. Our method is applicable to networks with static or dynamically allocated spectrum. Furthermore, we propose two algorithms for unique identification of the set of compromised nodes, one for independently acting nodes and one for colluding nodes. Detailed theoretical evaluation of the security metrics and extensive simulation results are provided to demonstrate the efficiency of our methods in mitigating jamming and identifying compromised nodes.