On the Data Gathering Capacity and Latency in Wireless

In this paper, we investigate the fundamental properties of data gathering in wirelesssensor networks, in terms of both transport capacity and latency. We consider a scenarioin which s(n) out of n total network nodes have to deliver data to a set of d(n) sink nodesat a constant rate f(n; s(n); d(n)). The goal is to characterize the maximum achievablerate, and the latency in data delivery. We present a simple data gathering scheme thatachieves asymptotically optimal data gathering capacity and latency with arbitrary net-work deployments when d(n) = 1, and for most scaling regimes of s(n) and d(n) whend(n) > 1 in case of square grid and random node deployments. Differently from mostprevious work, our results and the presented data gathering scheme do not sacrifice en-ergy efficiency to the need of maximizing capacity and minimizing latency. Finally, weconsider the effects of a simple form of data aggregation on data gathering performance,and show that capacity can be increased of a factor f(n) with respect to the case of nodata aggregation, where f(n) is the node density. To the best of our knowledge, theones presented in this paper are the first results showing that asymptotically optimal datagathering capacity and latency can be achieved in arbitrary networks in an energy efficientway