Detecting Misused Keys in Wireless Sensor Networks

Key management is the cornerstone for secure communication in sensor networks. Researchers have recently developed many techniques to setup pairwise keys between sensor nodes. However, these techniques allow an attacker to compromise a few sensor nodes and learn many pairwise keys used between non-compromised nodes. The attacker can then use these keys to impersonate non-compromised nodes and mislead the sensing application. To deal with this problem, this paper proposes to detect misused keys in sensor networks. The paper introduces a hidden layer of protection, which is designed for the security of pairwise keys rather than the messages in the network. It cannot be seen and will not be used by sensor nodes during normal communication. However, it can be checked by some special nodes to identify suspicious keys. With this idea, this paper develops a serial of techniques to detect misused keys. These techniques make it particularly difficult for an attacker to actively mislead the application using the compromised keys shared between non-compromised nodes. The paper also shows the effectiveness and efficiency of these techniques through analysis and experiments.