Received signal strength based indoor positioning algorithm using advanced clustering and kernel ridge regression

Le, Yanfen; Zhang, Hena; Shi, Wei; Yao, Heng

doi:10.1631/fitee.2000093

Cited by 10 publications

(2 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yang et al [33] proposed a WKNN indoor location algorithm based on spatial characteristics partition and localization restriction. Le et al [34] proposed an Advanced Clustering (AC) strategy approximating location by clustering matching and achieving precise location by kernel ridge regression. Wu et al [2] and Xu et al [35] proposed ViVi and ViViplus, respectively, and their key idea is to exploit the spatial awareness of RSS values by formulating FSG profiles or RSG matrices as enhanced WiFi fingerprints.…”

Section: Localization Methods Based On Wifimentioning

confidence: 99%

From Coarse to Fine: Two-Stage Indoor Localization with Multisensor Fusion

Zhang

Bao

et al. 2023

Tsinghua Sci. Technol.

View full text Add to dashboard Cite

Increasing attention has been paid to high-precision indoor localization in dense urban and indoor environments. Previous studies have shown single indoor localization methods based on WiFi fingerprints, surveillance cameras or Pedestrian Dead Reckoning (PDR) are restricted by low accuracy, limited tracking region, and accumulative error, etc., and some defects can be resolved with more labor costs or special scenes. However, requesting more additional information and extra user constraints is costly and rarely applicable. In this paper, a two-stage indoor localization system is presented, integrating WiFi fingerprints, the vision of surveillance cameras, and PDR (the system abbreviated as iWVP). A coarse location using WiFi fingerprints is done advanced, and then an accurate location by fusing data from surveillance cameras and the IMU sensors is obtained. iWVP uses a matching algorithm based on motion sequences to confirm the identity of pedestrians, enhancing output accuracy and avoiding corresponding drawbacks of each subsystem. The experimental results show that the iWVP achieves high accuracy with an average position error of 4.61 cm, which can effectively track pedestrians in multiple regions in complex and dynamic indoor environments.

show abstract

Section: Localization Methods Based On Wifimentioning

confidence: 99%

From Coarse to Fine: Two-Stage Indoor Localization with Multisensor Fusion

Zhang

Bao

et al. 2023

Tsinghua Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…In [50], AFP clustering is used at first and then the target ′ s location is estimated based on a hybrid distance which is a combination of the physical distance and the corresponding signal distance between two RPs. The fingerprint system proposed in [51] uses K-medoids clustering to create a set of overlapped clusters, combines the hamming distance between AP coverage vector with the Euclidean distance between the RSS vectors and then apply KRR method in the positioning phase. Such positioning system incurs high computational overhead.…”

Section: Related Workmentioning

confidence: 99%