Distortion, Energy and Capacity Aware Activation of Sensor Nodes in Wireless Sensor Networks
Izhak Rubin and Xiaolong Huang
We consider a sensor network involving sensors that are placed in specific locations. A point phenomenon is being detected and tracked by activated sensors. The latter collect data characterizing parameters of the phenomenon, possibly compress it and transport the compressed data to a central node. The latter processes the received data to derive an estimate of the phenomenon’s parameters. The corruption level of the sensed data is assumed to be dependent on the distance between the sensor and the point phenomenon. We also account for location dependent correlations between observation noise components. We develop computationally efficient algorithms for determining the specific set of sensors to be activated, and the level of compression to be applied by each selected sensor, under energy and communications capacity constraints. The aim is to achieve a sufficiently low reproduction distortion level. We demonstrate the performance behavior of our algorithms. We show that the activation of sensors that belong to a critical set of sensors tends to provide a distinct reduction in the distortion measure. In turn, the activation of additional sensors that are not members of this set may not lead to further distinct improvement.