UWB positioning for wireless embedded networks

The paper reports on the development of a low-power and low-cost device for low data rate communications with tracking and positioning capabilities. We investigate the performance of a range of position estimation methods making use of the estimate of the time-of-arrival (TOA) of the ultra wideband (UWB) signal at a set of receivers/sensors. The performance evaluation is performed in terms of the root-mean-square (RMS) error of the position coordinates estimation and the failure rate. We first discuss TOA estimation. A two-stage acquisition method is employed to speed up synchronization so to obtain the TOA estimate. We then study several optimization-based estimation methods including the Gauss-Newton type methods and the quasi-Newton method. Those methods are quite practical due to the fact that they do not require the knowledge of the characteristics or the TOA estimation error. Performance of the methods is examined through simulation. Furthermore, velocity of the moving devices and/or position estimate filtering can be exploited to improve the accuracy of the position estimation.