An adaptive scheme for data collection and aggregation in periodic sensor networks

Lowdata delivery efficiency and high energy consumption are inherent problems in sensor networks. Finding accurate data is difficult and leads to expensive sensor readings. Adaptive data collection and aggregation techniques provide opportunities for reducing the energy consumption of continuous sensor data collection. In this paper, we propose an adaptive data collection and aggregation scheme for periodic sensor networks. Our approach is two-fold: firstly, we present an efficient adaptive sampling approach based on the dependence of conditional variance on measurements varies over time. Then, over-sampling can be minimised and power efficiency can be improved. Furthermore, we propose a multiple levels activity model that uses behaviour functions modelled by modified Bezier curves to define application classes. Secondly, a new data aggregation method is applied on the set of data collected from all sensor nodes. We investigate the problem of finding all pairs of nodes generating similar datasets. Our proposed models are validated via simulation on real sensor data.