Research On The Algorithm Of TOA Estimation In IR-UWB Ranging System

Impulse radio ultra-wideband (IR-UWB) is a promising pulse-based carrierless communication technology. With the advantages of the high positional accuracy, great anti-interference performance, high data communication rates, and low power consumption, it becomes one of the first choices of technology for future communication and position estimation applications. Nowadays, most of position estimation methods are mainly based on the time of arrival (TOA) estimation of the direct path (DP) between the transmitter and the receiver which is however often affected by indoor noise, multi-path propagation, and non-light-of-sight (NLOS) environments. Therefore, it is a great challenge of IR-UWB ranging technology to achieve higher estimation accuracy in the multi-path and NLOS scenarios.In recent years, the research on TOA estimation algorithms in IR-UWB systems can be roughly divided into two categories: the coherent algorithm based on matched filter detection and non-coherent algorithm based on energy detection. It has been recommended by various standard criteria proposed by IEEE802.15.4a that energy detection scheme is suitable for ranging and positioning with a low rate and low power consumption. Therefore, the study of non-coherent TOA estimation based on energy detection attracts many attentions of the domestic and foreign scholars.The main objective of this dissertation is to reduce the influence of noise and multipath problems and improve the ranging accuracy in IR-UWB ranging systems. Firstly, we introduce the characteristic of UWB signal and channel propagation model, and analyze the path loss, multipath delay and other effects on ranging accuracy which provide the mathematical model for the discussion of TOA estimation in the following chapters. Secondly, the advantages of ranging with UWB are analyzed according to the characteristics of UWB. We then introduce several traditional TOA algorithms based on TOA ranging technology and analyze their lower bounds. Thirdly, we propose a sawtooth-correlated TOA estimation algorithm based on the existing non-coherent energy detection algorithm. We extract the envelope information of the received signal comes from the output of energy detector, then consider the sawtooth waveform as a local template signal which is correlated to the received signal, finally, we can get the TOA estimation of DP by evaluating the relevant peak positions. Simulation results show that, with the advantage of simple and precise estimation, the proposed algorithm not only omits the process of threshold setting, but also reaches the similar estimation accuracy as the TC algorithm. However, the traditional UWB TOA estimator has the limitations of high complexity and poor performance in the low SNR region. We propose a least squares (LS) approach with wavelet de-noising which improves the estimation accuracy at low SNR. Our algorithm estimates the TOA as a by-product of the LS channel estimator based on a threshold technique, which is simple and can enable fast, accurate TOA estimation applicable to real-time applications. In the simulation result, we demonstrate that the proposed method can achieve great robustness with high ranging accuracy at low SNR region, where the estimation performance of TOA under CFA standard outperforms the fixed threshold.