Open AccessAdaptive indoor positioning method based on direction discrimination and device conversion
Author(s) -
Li Shirong,
Fu Maosheng,
Zhu Xuemei,
Zhang Fenghui,
He Fugui
Publication year - 2020
Publication title -
iet wireless sensor systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.433
H-Index - 27
eISSN - 2043-6394
pISSN - 2043-6386
DOI - 10.1049/iet-wss.2019.0079
Subject(s) - rss , computer science , piecewise , curse of dimensionality , mobile device , least squares function approximation , workload , principal component analysis , artificial intelligence , algorithm , mathematics , statistics , mathematical analysis , estimator , operating system
Received signal strength (RSS) greatly differs due to the different occlusion directions and receiving device heterogeneity. It greatly affects the positioning accuracy. In this study, an adaptive indoor positioning method based on the direction discrimination and device conversion is proposed to solve these problems. This method is mainly composed of three parts: direction discrimination, device conversion and positioning models. First, the direction discrimination model can reduce the impact of a user's body occlusion. Best access points can be selected by principal component analysis to adapt to different directions and areas. Secondly, a device conversion model is used to reduce high offline work due to device heterogeneity. RSS of other devices can be converted to the value of one fixed device by least squares piecewise polynomial algorithm, without increasing the offline data collection workload. Finally, the results can be obtained by the positioning model. The problems of high dimensionality and non‐linearity can be solved by the least squares support vector regression algorithm. Experimental results show that the proposed method can solve the problems of occlusion direction and device heterogeneity. The engineering applicability of positioning system can also be greatly improved.