Bad Geometry Influence on Positioning Accuracy in Wireless Networks

The aim of this paper is to discuss a connection between the geometric configuration and positioning accuracy in wireless networks exploiting the TDOA technique. At the beginning, a simulation model of TDOA localization technique is described. We use the Taylor series expansion and least square fit method to estimate the terminal's position. In this part we also derive a statistical model for calculating the geometric dilution of precision (GDOP), which is here a variable reflecting geometric quality with respect to the hyperbolic lateration. The two dimensional root mean square (RMS) error metric is also presented in order to make necessary accuracy calculations. We then describe an impact of the bad configuration of measurement units on the positioning accuracy. Basing on simulation results, the TDOA efficiency for different time errors and geometric circumstances is presented. Finally, we show an empirical solution for decreasing the bad geometry effect. The accuracy is improved thanks to the additional measurement units placed in the good geometry conditions.