The University of British Columbia

We design transmission schemes in cooperative MIMO sensor networks that optimize the choice of rate and cluster for minimizing the energy consumption subject to a given average probability of error. In general it was observed that, the number of clusters to optimize the energy consumption is a non-increasing function of long haul distance. By virtue of this, we conclude that as the long haul distance increases the cooperation among the nodes increases. Also, for large long haul distances the optimal number of clusters is shown to converge to a constant. Next, we propose a provably convergent block coordinate descent algorithm to determine the optimal joint rate and number of clusters. Through our numerical results it was observed that a cluster optimized cooperative MIMO transmission scheme can be more energy efficient than a rate only optimized scheme. Also a joint rate and cluster optimized transmission scheme can yield large scale energy savings for short and medium range applications compared to the rate only or cluster only optimized transmission schemes.