A New Method to Discriminate Wireless Location under Noisy Circumstance

Location distinction, as the ability of a receiver to determine whether a transmitter has changed its position, is important for many wireless applications. In order to improve the performance of location distinction in noisy environment, we propose a novel method based on ordinal pattern measure. The existing methods, such as channel gains of multi-tonal probes, received signals strength and temporal impulse response, take rare consideration on the impact of noise at the receiver. We evaluate the above methods in the noisy environment and find dramatic performance degradation in low SNR situation. We use a 2.4 GHz link measure data set, obtained from CRAWDAD, and find that the ordinal pattern method significantly reduces the false alarm rate in the noisy environment, in comparison to the original temporal impulse response method.