UWB positioning using six-port technology and a learning machine

This paper presents a short-range positioning system based on six-port technology and the corresponding signal processing algorithms. Accurate positioning with high resolution of detected targets is achieved by utilizing both the impulse signal and the wideband phase discrimination characteristic of a six-port circuit operating in ultra-wideband (UWB) mode. This system has several advantages such as object penetration, multipath immunity and low probability of interception. Signal processing includes range estimation and position location algorithms. The range is estimated by analyzing the output signal pattern with support vector machines (SVM). Based on the estimated ranges between a target node and access points, the position of the target is determined using hyperbolic position location method. Ranging and position simulation results of the considered system are given for an indoor multipath channel. Accurate positioning has been achieved with an average root mean square error of 0.68 cm. The simulation results also show the robustness of the proposed system in a multipath channel