Detecting substitution attacks against non-colluding relays

The goal of this paper is to obtain the channel conditions (if exist) under which substitution attacks performed by relay node(s) in a relay network can be detected. The network model considered consists of a source node and a destination node. There are two independent transmission paths from the source to the destination, each via a potentially malicious relay which may perform substitution attacks by forwarding altered symbols to the destination. The destination attempts to detect any such malicious act of the relays by comparing the joint empirical distribution of the symbols received from the relays with known channel statistics along the two paths. Note that every symbol received by the destination may be altered, and hence no clean reference observation is available to the node. It is demonstrated that maliciousness of the relays can be asymptotically detected with sufficient channel observations if and only if the two relays do not collude and the network satisfies a non-manipulability condition.