期刊名称:International Journal of Distributed Sensor Networks
印刷版ISSN:1550-1329
电子版ISSN:1550-1477
出版年度:2019
卷号:15
期号:7
页码:1
DOI:10.1177/1550147719866135
出版社:Hindawi Publishing Corporation
摘要:In indoor positioning techniques, Wi-Fi is one of the most used technology because of its availability and cost-effectiveness. Access points are usually the main source of Wi-Fi signals in an indoor environment. If access points are optimized to cover the indoor area, this could improve Wi-Fi signal distribution. This article proposed an alternative to optimizing access point placement and distribution by introducing virtual access points that can be virtually placed in any part of the indoor environment without installation of actual access points. Virtual access points will be created heuristically by correlating received signal strength indicator of already existing access points and through linear regression. After introducing virtual access points in the indoor environment, next will be the addition of filters to improve signal fluctuation and reduce noise interference. Kalman filter has been previously used together with virtual access point and showed improvement by decreasing error distance of Wi-Fi fingerprinting results. This article also aims to include particle filter in the system to further improve localization and test its effectiveness when paired with Kalman filter. The performance testing of the algorithm in different indoor environments resulted in 3.18 and 3.59 m error distances. An improvement was added on the system by using relative distances instead of received signal strength indicator values in distance estimation and gave an error distance average of 1.85 m.
关键词:Wi-Fi; received signal strength indicator; fingerprinting; localization; virtual access point; Kalman filter; particle filter
其他关键词:Wi-Fi ; received signal strength indicator ; fingerprinting ; localization ; virtual access point ; Kalman filter ; particle filter