Time-of-Arrival Estimation for Positioning in Bandwidth-Limited Dense Multipath Channels

Abstract

For time-of-flight-based wireless positioning systems operating in (dense) multipath propagation channels, the accuracy is severely influenced by the signal bandwidth, because the dense multipath component (DMC) interferes with the desired, information-bearing line-of-sight (LoS) signal. Several such systems make use of bandwidth-limited frequency resources, e.g. the industrial, scientific and medical (ISM) bands, therefore the achievable position estimation accuracy is limited. In this paper, we propose a model-based delay-estimation method which takes into account a parametric model of the DMC and thus exploits the signal energy carried in the DMC. The resulting algorithm exhibits an enhanced delay estimation accuracy and remarkable robustness in non-LoS situations. The algorithm is benchmarked against a maximum likelihood (ML) estimator not incorporating a model for the DMC and against the estimated Cramér-Rao lower bound (CRLB) in presence of DMC. Results show a significant performance gain for scenarios where the conventional ML estimator performs poorly. An evaluation of measurement data validates the simulation, showing a root-mean-square error (RMSE) of 33.4 cm compared to 1.89 m for the conventional ML estimator, at a signal bandwidth of 80 MHz.