Open AccessSAPIL: single access point based indoor localisation using Wi‐Fi L‐shaped antenna array
Author(s) -
Tian Zengshan,
Lian Yinghui,
Zhou Mu,
Pu Qiaolin
Publication year - 2019
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.2018.5129
Subject(s) - computer science , multipath propagation , non line of sight propagation , impulse (physics) , algorithm , multilateration , angle of arrival , antenna (radio) , real time computing , signal (programming language) , path (computing) , line of sight , channel (broadcasting) , acoustics , telecommunications , wireless , physics , computer network , quantum mechanics , node (physics) , astrophysics , programming language
Single access point (AP) based indoor localisation (SAPIL) has attracted much attention recently. Most of the existing SAPIL systems require frequency hopping, which cannot be easily achieved in commercial APs. In response to this compelling problem, the authors proposed the SAPIL using Wi‐Fi L‐shaped antenna array without modifying the existing Wi‐Fi protocol. Specifically, the scenario in which the target is located is determined based on the maximum weight firstly. If the target is located in the line‐of‐sight scenario, the angle‐of‐arrival (AoA) corresponding to direct path is identified from the AoAs of multipath signals which are estimated using the smooth multiple signal classification algorithm. Meanwhile, the adaptive signal propagation model trained from the channel impulse response in offline phase is adopted to calculate the distance of direct path. Then the target is located by the estimated AoA and distance of direct path. Otherwise the target is located by combining its speed and heading with pedestrian dead reckoning. Finally, the experimental results demonstrate that the proposed system achieves the median AoA estimation error of 5°, median distance estimation error of 0.5 m, and median localisation error of 1.8 m.