CFAR-Based TOA Estimation and Node Localization Method for UWB Wireless Sensor Networks in Weibull Noise and Dense Multipath

In the presence of dense multipath, the threshold-based time of arrival (TOA) estimation methods have attracted interest for node localization in ultra-wideband (UWB) wireless sensor networks (WSNs). Since the first arrival path is not always the strongest one in many actual situations, we propose to determine the real TOA using the constant false alarm rate (CFAR) detector method. Unlike conventional Gaussian or Rayleigh background, we assume the amplitude of the observal noise to be Weibull distributed. Given the probability of false alarm, the threshold of TOA detector is derived based on Weibull distribution function. Using the obtained TOA values, the multilateral localization is employed to compute the positions of nodes. The three-dimensional (3D) Taylor algorithm is given to solve the non-linear equations. The simulations prove that the proposed method has high time-resolution and low computational load in positioning of WSN nodes under Weibull noise and dense multipath.