Utilizing Kriging to Generate a NLOS Error Correction Map for Network Based Mobile Positioning

Abstract

Network mobilephone-based positioning experiences degradation of location accuracy due to localised non-line-of-sight (NLOS) signal propagation. This is well known to be a major source of error in network-based mobilephone positioning. NLOS error systematically causes the Mobile Station (MS) to appear further away from the base station than it actually is, thereby increasing the positioning error. One method to mitigate the effect of NLOS error is to generate a NLOS error correction map, and then use the correction map to rectify the distorted MS location. The correction map can be generated using the following procedure: (1) estimating the NLOS errors at points where the real positions can be obtained utilising other information such as the points very near BTS (Base Transceiver Station) and the intersections of streets, or the location where the measurement has been made; and (2) interpolating or extrapolating the errors to specific points that we are interested in. Assuming some reference points have been obtained, this paper utilises kriging, an estimation technique that is widely used in mining, to generate the correction map. Theoretically kriging can also be used wherever a continuous measure is made on a sample at a particular location in space or time. Using simulations with a typical dense urban environment assumption, the feature of the NLOS error variogram is analysed and different models of the variogram are compared. The correction map of NLOS error is generated using some 'sampled' points, and compared with the 'true' NLOS error map to show the efficiency of kriging.