A practical UWB TOA estimator in the presence of interference and NLOS

Ultra-wideband (UWB) technology, using time-of-arrival (TOA)-based ranging techniques, is a viable candidate for future location-aware sensor networks. In this paper, we propose an optimized threshold selection technique for UWB TOA estimation. The proposed approach derives the threshold by applying the linear regression to the TOA estimation error probability surface and is applicable to the real-time systems without any prior information. We show that the proposed technique outperforms many other well-known techniques in most SNR regions through the implementation in CM4 channel model. We also demonstrate its superiority to other techniques in the presence of the interference and NLOS.