Experimental verification of bias reduction in TDOA based localization

Abstract

In this paper, we verify a bias reduction algorithm proposed earlier through experiments. The accuracy of localization systems are influenced by many factors. Our focus is to improve the performance of localization systems by reducing the estimation bias. Many algorithms have been proposed to reduce bias, however, only simulation results are presented which only demonstrate the algorithms with ideal or generated synthetic data, so their performance in practical situation is unknown. To address this problem we conducted real-world experiments. The experimental data is obtained from a TDOA based localization system developed using Software Defined Radios. The results demonstrate that the proposed bias reduction algorithm can reduce the bias effectively, thus greatly enhancing the localization accuracy of the system.