Space-time coding for distributed detection in wireless sensor networks

We consider the problem of distributed detection in wireless sensor networks, where sensors send their local binary decisions over error-prone wireless channels, subject to noise and fading, to a fusion center (FC) that is tasked with making the final decision. This is different from classical literature on distributed detection which provides decision and data fusion strategies assuming that sensors´ local decisions are sent over orthogonal error-free channels, without taking into account fading and noise. To mitigate the deteriorating effect of fading, we propose to employ space-time coding (STC) at sensors for transmission of local decisions to the FC. We consider a communication architecture, in which the sensors are grouped into pairs, and the sensors in each pair communicate with the FC, after exchanging their binary decisions. This architecture allows us to utilize Alamouti´s space-time coding (STC) at the sensors. The advantage of this scheme is to combat the fading effect via harvesting diversity and thus to improve significantly the detection reliability in the presence of fading. We derive the optimal Likelihood ratio test (LRT) fusion rule and study the effect of communication errors when sensors are exchanging decisions over inter-sensor channels on the overall detection performance. Numerical results confirm the effectiveness of the proposed scheme, and make STC a suitable candidate for signaling in distributed detection systems operating in a fading environment.