Two Novel Methods for Accurate NLOS Detection Based on Channel Statistics

Time-of-arrival (TOA) estimation is the first step of the most positioning algorithms. However in various environments especially when ultra wideband (UWB) pulses are used, TOA extraction from the received signal is challenging. UWB radio propagation bears multipath phenomenon, therefore correct identification of the first path TOA highly depends on the statistical characteristics of the environment and apprehension that the signal has been passed through the line-of-sight (LOS) channel or the non-line-of-sight (NLOS) one. In this paper, two novel NLOS identification techniques based on the multipath channel statistics are proposed. Simulations show that the first technique using the fitness equations of mean and variance of the received signal is suitable for residential and outdoor environments. The other one that compares the relative energy of two different periods of the received signal is more accurate in office and industrial environments. IEEE 802.15.4a channel models are used and two hypothesis tests are applied to distinguish between LOS and NLOS. The high accuracy identification of channel type is achieved for all mentioned environments. KEYWORDSUWB; channel; TOA; Ranging; optimization; IEEE 802.15.4a