Scattering Point Estimation in UWB Systems Using a Grid-Based Filter

Abstract

The localization of Vulnerable Road Users (VRUs) in traffic environments is critical for ensuring safety, particularly for passive users who are not equipped with communication devices. Multipath-enhanced Device-Free Localization (MDFL) has emerged as a promising method for VRU positioning by exploiting multipath components (MPCs). However, MDFL relies on accurate knowledge of the static scattering and reflection points that generate these MPCs, which is often unavailable in dynamic traffic scenarios. In this work, we propose a grid-based filter to estimate the locations of static scattering points online in real traffic environments. Our approach utilizes the amplitudes of MPCs and correlates induced signal attenuations to the positions of tracked objects. We validate the proposed method through simulations, demonstrating that the scattering point locations can be accurately estimated within two minutes, depending on the noise level, the number of tracked objects, and their positional accuracy. Additionally, we validate the approach in real traffic experiments, successfully pinpointing an unknown reflection point, highlighting its practical applicability and robustness.