Scaling laws and operation of wireless ad-hoc and sensor networks

We present a novel approach based on coherent beamforming for increasing throughput in a wireless ad-hoc network. These results follow Gupta and Kumar's 'optimal' throughput capacity results, which was shown to follow a scaling law of O(1//spl radic/n) bits/sec per node. The result therein is based on a communication model that is primarily characterized by transmit power resulting in omni-directional transmission. Our communication model is characterized by both phase and transmit power. Consequently, with cooperation among several neighboring nodes we can achieve directional transmission and reception by means of coherent beamforming. We construct a feasible routing scheme and show that a throughput capacity of O (W/n/sup 1/3/) bits/sec per node is achievable, where W is the link capacity and n nodes are uniformly and randomly distributed in a unit area. Although, these bounds still imply vanishing throughput capacity per node, the improved scaling law has implications for the problem of estimation of a spatially correlated random field to an arbitrary degree of accuracy in distributed sensor networks.