Positioning in Wireless Sensor Networks Using Array Processing

In this paper, a novel positioning approach for wireless sensor networks (WSNs) is proposed. The approach locates a given target node by using a number of surrounding anchor nodes as an array of large aperture with known geometry. The approach uses the concept of the array manifold vector based on spherical wave propagation. The position of the target is estimated as the intersection point of three circular loci, the centres and radii of which are calculated from the signal eigenvalues of the covariance matrices corresponding to a number of reference points. A number of simulation results show that the proposed approach can estimate a target location with exceptional accuracy even for low SNR_in, x (observation-interval). Analytical expressions are also provided for the estimation error due to the finite averaging effects, as a function of the locations of the anchor nodes and the SNR_in.