Adaptive Anchor Deployment for Diverse Demands on Localization Precision

Precise localization technology promotes many modern applications, including automatic driving, navigation and so on. The ultra-wideband technology (UWB) is a promising localization method due to its high-precision and low power consumption. During its localization, a proper anchor deployment can not only improve localization accuracy, but also enhance resistance to interferences. In the current researches, the deployment of UWB anchor nodes is optimized by minimizing the Cramer-Rao lower bound (CRLB). However, it ignores non-line-of-sight (NLoS) interference caused by fixed obstacles and high-precision requirements in a specific region. Hence, we propose a comprehensive solution to search for the optimal deployment of UWB anchors under LoS or mixed LoS/NLoS propagations, and consider the diverse demands on UWB localization precision. Firstly, the integral of the localization mean square errors (MSEs) is derived by Gaussian approximation in mixed LoS/NLoS scenarios. Secondly, region of interest (ROI) and dimension of interest (DOI) are further taken into account and the localization performance with respect to anchor deployment is presented numerically. Finally, a comprehensive particle swarm optimization (PSO)-based optimal anchor deployment searching algorithm is proposed. The simulation results show that the proposed optimal deployment of anchors outperforms three other typical anchor deployment schemes. Furthermore, the effectiveness of the proposed anchor deployment optimization is verified in our UWB test platform. Compared with three other deployments, its localization performance is improved by 83.3%, 41.8% and 40.7%, in terms of root mean square error (RMSE) in the mixed LoS/NLoS scenarios with a fixed obstacle.