Routing and aggregation strategies for contour map applications in sensor networks

Abstract

We consider the limitations inherent in data aggregation techniques for wireless sensor networks. In particular, we illustrate the limitations of schemes in which data from a sensor are routed to the node that can perform the most effective data aggregation. We then seek alternative forms of aggregation where, in principle, it is possible to aggregate readings from many sensors into a single one. To this end, we select contour maps as the most convenient example of data representation because they are used by various applications and, at the same time, allow for aggressive data aggregation. We describe particular aggregation and routing strategies that exploit the structure of contour maps. Namely, we use clustering based on the contour map values, as well as intra-cluster and inter-cluster routing and aggregation heuristics. Simulation results demonstrate the capabilities and limitations of the various varieties of aggregation/routing strategies. © 2009 Springer Berlin Heidelberg.