On the Data Gathering Capacity and Latency in Wireless Sensor Networks

In this paper, we investigate the fundamental properties of data gathering in wireless sensor networks, in terms of both capacity and latency. We consider a scenario in which s(n) out of n total network nodes have to deliver data to a set of d(n) sink nodes at a constant rate lambda(n, s(n), d(n)). The goal is to characterize the maximum achievable rate, and the latency in data delivery. We present a simple data gathering scheme that achieves asymptotically optimal data gathering capacity and latency with arbitrary network deployments when d(n) = 1, and for most scaling regimes of s(n) and d(n) when d(n) > 1 in case of square grid and random node deployments. Differently from most previous work, our results and the presented data gathering scheme do not sacrifice energy efficiency to the need of maximizing capacity and minimizing latency. Finally, we consider the effects of a simple form of data aggregation on data gathering performance, and show that capacity can be increased by a factor f(n) with respect to the case of no data aggregation, where f(n) is the node density. To the best of our knowledge, the ones presented in this paper are the first results showing that asymptotically optimal data gathering capacity and latency can be achieved in arbitrary networks in an energy efficient way.