Robust Short-Delay Multipath Estimation in Dynamic Indoor Environments for 5G Positioning

In urban and indoor settings, the efficacy of the global navigation satellite system is notably limited, prompting a shift towards utilizing cellular and wireless signals for location services. However, existing methods struggle to discern short-delay multipath signals in intricate indoor environments, often faltering in the presence of non-Gaussian noise. This paper introduces the Recursive Maximum Correntropy Criterion based Short-Delay Multipath Estimation (RMCSME) algorithm as a solution. By leveraging a short-delay multipath signal processing model and the recursive correntropy criterion, RMCSME accurately estimates dynamic multipath signals in the presense of non-Gaussian noise challenges. Through simulations and empirical signal tests, RMCSME demonstrates a marked reduction in ranging errors attributable to multipath effects while maintaining computational efficiency. Comparative analyses with the Improved Multipath Estimation Delay-Locked Loop (IMEDLL) and Multiple Signal Classification (MUSIC) algorithms reveal that the RMCSME algorithm performs better in static experiments. In dynamic tests, RMCSME achieves a positioning accuracy of 0.29 meters, surpassing IMEDLL by 21.7% and MUSIC by 39.6%. Furthermore, this approach presents a novel strategy for mitigating short-delay multipath errors in indoor 5G positioning signals, providing crucial support for achieving precise localization in commercial 5G networks.