Ranging Error Compensation Algorithm for Vehicle Key Localization in NLOS Environments

With the widespread adoption and development of smart vehicles, using the location of vehicle key with short-range signals, such as Bluetooth to achieve door unlocking, locking, and engine start have garnered significant attention. However, in the processing of vehicle key localization with distance measurement, the line-of-sight (LOS) and non-line-of-sight (NLOS) conditions between the vehicle key and the vehicle may frequently switch due to occlusions by obstacles such as the human body. Traditional NLOS ranging error compensation algorithms struggle to achieve satisfactory compensation results under uncertain movement states of the vehicle key due to low recognition accuracy and the single-motion model of Kalman filtering. To address this problem, this paper designs a ranging error compensation algorithm based on the Interacting Multiple Model (IMM), which resolves the issue of the single-motion model in traditional Kalman filtering algorithms. Then, using the Wireless Insite simulation software, the human movement trajectory in a vehicle key scenario is modeled. Finally, the results show that the IMM-based ranging error compensation algorithm provides more significant error suppression and compensation for NLOS-induced errors compared to the single-model Kalman filtering algorithm, leading to a greater improvement in real-time ranging accuracy.