Research on indoor high accuracy positioning method based on dynamic weighted fusion of 5G-RSS and TDOA

Aiming at the problems of low single positioning accuracy and insufficient stability in complex indoor environments, this paper proposes a high-precision positioning method based on dynamic weighted fusion of 5G received signal strength (RSS) and time difference of arrival (TDOA). The signal quality is quantified by defining environmental factors K, the RSS and TDOA weights are adjusted in real time, and a nonlinear joint error function is constructed for gradient descent optimization. The experiments show that: in low-quality environment (K < 1), dynamically boosting the TDOA weights to 60%, the average error and standard deviation of positioning are reduced by 33.7% and 37.7%, respectively, compared with the fixed-weight 6:4 method; high-quality scenario (K ≥ 1), with more than 70% of the RSS weights, the 5-node fusion error (2.63m) approaches the accuracy of a single TDOA at 10 nodes (2.28 m). The method significantly reduces the error volatility under node sparsity and dynamic interference (standard deviation optimization of 31%-48%), providing a more robust technical solution for indoor high-precision positioning.