Distributed detection of replicas with deployment knowledge in wireless sensor networks

A number of protocols have been developed to mitigate the threat against wireless sensor networks of an attacker finding nodes, compromising them, and using these nodes to eavesdrop or undermine the operation of the network. A more dangerous threat that has received less attention, however, is that of replica nodes, in which the attacker compromises a node, extracts its keying materials, and produces a large number of replicas that he can spread throughout the network. This attack enables the attacker to leverage the compromise of a single node to create widespread effects on the network. To defend against these attacks, we propose distributed detection mechanisms to find and revoke replicas. Our schemes are based on the assumption that nodes are deployed in groups, which is realistic for many deployment scenarios. By taking advantage of group deployment knowledge, our schemes perform replica detection in a distributed, efficient, and secure manner. We show analytically and in simulation that our schemes achieve effective and robust replica detection capability with substantially lower communication, computational, and storage overheads than prior work.