This letter presents the variation of localization error to network parameters, the number of range estimation results from anchor nodes (ANs) and average distance between ANs in centralized Wireless Sensor Network (WSN). In sensor network, ANs estimate the relative range to Target Node (TN) using Time-Of-Arrival (TOA) information of Ultra WideBand (UWB) radio and a fusion center determines the final localization of TN based on estimation results reported. From simulation results, the variation of localization error, which is defined as the difference between localization result of TN and its actual location, is represented as the function of number of estimation results to average distance between ANs. The distribution of localization error is matched to the Rician distribution whose K-factor value is given by the proposed formula as well. Finally, the normalized error function for the efficient localization network design is characterized.

This letter proposes an Ultra WideBand (UWB) channel gain estimation algorithm that is based on frequency response correlation. From data measured in indoor and outdoor environments, the frequency correlation statistics of the UWB channel are characterized by correlation coefficients, which are represented as functions of the frequency separation according to the propagation environments. We use empirically established correlation coefficient models to develop the proposed channel gain estimation algorithm. In this algorithm, unknown channel gains of intermediate frequencies are estimated from known channel gains of neighboring reference frequencies. The performance of the proposed algorithm is evaluated against various configurations of reference frequencies and compared with that of a conventional estimation algorithm using a linear interpolation scheme.