Joint TOA/RSS Positioning in Indoor NLOS Environments for Cellular Networks

Cellular-based positioning has garnered significant attention for indoor localization, particularly in non-line-of-sight (NLOS) environments where conventional techniques exhibit severe accuracy degradation. This article proposes a joint time of arrival (TOA) and received signal strength (RSS) approach to enhance positioning performance in indoor NLOS scenarios. To mitigate NLOS errors, we develop an alternating residual compensation (ARC) algorithm that iteratively compensates for measurement biases. Furthermore, leveraging sparse anchor points (APs), we construct a TOA/RSS environmental map using deep Gaussian processes (DGPs) to provide supplementary propagation characteristics in complex indoor layouts. Extensive simulations and real-world experiments validate the proposed method. Simulations demonstrate submeter positioning accuracy, while real-world tests in operational 5G environments achieve an average error below 1.5 m. These results demonstrate the robustness of the proposed approach, making it a promising solution for cellular networks in indoor NLOS Environments.