A fast data aggregation technique for wireless sensor networks

A wireless sensor network is composed of a large number of sensor nodes embedded in the physical world to perform monitoring and surveillance tasks. The primary operation on sensor networks is extracting aggregated information from the networks, which can be time-consuming due to the environmental dynamics and the large number of sensor nodes involved. In this thesis, we present a fast data aggregation technique for wireless sensor networks which uses a randomized architecture. The architecture is composed of layers which are constructed in a distributed, localized fashion. The key property of our technique is that data are collected from one layer containing a subset of sensor nodes, resulting in fewer hops and thus lower delay in data aggregation. We provide theoretical guarantees for the delay incurred and the accuracy level of the results. In addition, we explore ways to further speed up the data aggregation process by using history information.