1. Map-aided and UWB-based anchor placement method in indoor localization
- Author
-
Xiaogang Qi, Hao Pan, Lifang Liu, and Meili Liu
- Subjects
0209 industrial biotechnology ,Computer science ,Heuristic (computer science) ,Ultra-wideband ,02 engineering and technology ,Evaluation function ,Multilateration ,Non-line-of-sight propagation ,020901 industrial engineering & automation ,Artificial Intelligence ,Differential evolution ,0202 electrical engineering, electronic engineering, information engineering ,Path loss ,020201 artificial intelligence & image processing ,Cramér–Rao bound ,Algorithm ,Software - Abstract
Nowadays, ultra wideband (UWB) has been popular in indoor positioning systems. Because of obstacles such as walls, doors, desks and pedestrians in the indoor area, UWB devices have to be conducted in a non-line-of-sight (NLOS) environment. The placement of nodes has significant influences on the performance of localization. In this paper, an anchor placement method for the target’s trajectory based on genetic heuristic differential evolution algorithm is proposed to improve the UWB localization accuracy. To be specific, in an area with low anchor densities, a target does not have enough anchors for multilateration localization; thus, a coverage degree criterion is defined. Meanwhile, the Cramer–Rao lower bound (CRLB) is used as an evaluation metric for localization accuracy, and both of CRLB and coverage degree criterion are incorporated into the evaluation function of the differential evolution (DE) algorithm. Furthermore, instead of using the liner distance path loss model, the more practical UWB-based through-wall signal propagation model is adopted to reflect the influences of obstacles that are widespread in indoor environments. In addition, a software application is developed to visualize the indoor scenario, signal propagation status and anchor placement results. Finally, field experiments and simulations are performed, and a thorough case study confirms the effectiveness of the proposed method. The average localization error of the proposed placement scheme can be reduced by 28.2% and 12.5% compared to the random scheme and the default DE-based scheme.
- Published
- 2021