Detecting Malicious Behavior in Cooperative Diversity

We consider a cooperative diversity scheme where a relay cooperatively enhances communication between a source and destination. In cooperative diversity, due to lack of a mechanism to ensure relay's adherence to the cooperation strategy, the receiver is often assumed to be passive. In this paper, we consider a smart destination which examines relay's signal prior to applying diversity combining. This is attributed to the assumption that relay may not conform to the cooperation strategies at all times and may behave maliciously. Based on this assumption, we develop a statistical detection technique to mitigate malicious relay behavior in decode-and-forward cooperation strategy. The detection technique statistically compares the signals received from the two diversity branches to determine the relay's behavior. As the uncertainty in the direct path is only due to the channel, correlation of received signals from the source and relay provides a basis to characterize relay behavior. We show, both by analysis and simulation, that a malicious relay reduces the correlation between the received signals in the diversity branch. Finally, we investigate bit-error rate and outage behavior performance in the presence of a smart destination.