Multisensor accuracy enhancement on the base of interval voting in form of preference aggregation in WSN for ecological monitoring

In wireless sensor networks (WSNs) for ecological monitoring, sensing performance degrades over time due to harsh environmental conditions or battery depletion. Calibration of multisensors requires extra temporal, financial or energy resources. In order to enhance an accuracy of the multisensor measurements, we propose an approach based on interval voting in form of preference aggregation. The approach allows to determine a corrected value of a measured parameter, lying within narrow uncertainty bounds, on the basis of inaccurate measurement data, obtained from neighboring multisensors. Simulation results have shown that the proposed approach allows to achieve considerably better accuracy comparatively to other known interval voting techniques.