LEAP: Localized Energy-Aware Prediction for data collection in wireless sensor networks

For many applications in wireless sensor networks, accurate data collection is a crucial problem. Users may want to continuously extract data from the networks for analysis after. Clustering and prediction techniques, which exploit spatial and temporal correlation among sensor data, provide opportunities for reducing the energy consumption of sensor data collection. We propose the LEAP (localized energy-aware prediction) approach. LEAP is clustering based. A cluster head represents all sensor nodes in the cluster, and collects data values from them. LEAP implements local prediction algorithms, and only data values not within a specified error bound are collected by a cluster head. By doing so, the cluster head maintains an accurate view of the sensor data, while the communication cost is reduced. In this paper, we present energy-aware prediction models used in LEAP, analyze the performance tradeoff between reducing communication cost and limiting prediction cost, and design algorithms to exploit the benefit of energy-aware prediction. We believe LEAP has broad applications. Our proposed models, analysis, and algorithms are validated via simulation.