Interpolation for wireless sensor network coverage

Abstract

One of the primary issues for any wireless sensor network (WSN) deployment is that of longevity of the network. The lifespan of the network must be maximized while maintaining the quality of the sensory data received from the network. Most current solutions for achieving this require the definition of a sensing radius for each sensor together with a coverage level k. The sensed area is covered if every point within the area is inside the sensing radius of at least k sensors. When some points are covered by more than k sensors, it may be possible to instruct the surplus sensors to enter a low power sleep mode. This will conserve the energy of the network while maintaining the required coverage level. In. this paper we propose a novel alternative to this approach based on interpolation. If the sensor network is capable of interpolating the sensed medium at a given sensor's location to a specified accuracy or higher, then we propose that this sensor is redundant and can be put into sleep mode. We demonstrate our approach using live sensory data.