ALTAR: Area-based Localization Techniques using AoA and RSS measures for Wireless Sensor Networks

Mamidi, Kiran Kumar; Prasad, K. N. R. Surya Vara

doi:10.1109/ic3i46837.2019.9055582

Cited by 2 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Extending the example above, Gaussian-based fingerprinting computes both the mean (µ i j ) and standard deviation (σ i j ) of RSSI measurements collected by sensor S j with regard to reference point P i . The Gaussian distribution is expressed in (3).…”

Section: Fingerprintingmentioning

confidence: 99%

“…High variability in RSSI measurements due to fading (path-loss), shadowing (temporary obstruction between a sender and a receiver), and interference (overlap of WiFi channels) is a significant challenge in RSSI-based localization [2]. To mitigate RSSI variability, some active localization methods leverage device cooperation to obtain additional location-sensitive information, such as signal angle of arrival [3,4], round-trip time [5], device orientation [6], and device prior location [7]. However, passive localization cannot benefit from these methods, as it does not have the luxury of device cooperation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mobility Intelligence: Machine Learning Methods for Received Signal Strength Indicator-based Passive Outdoor Localization

Bao¹,

Mazokha²,

Hallstrom³

2022

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

Knowledge of pedestrian and vehicle movement patterns can provide valuable insights for city planning. Such knowledge can be acquired via passive outdoor localization of WiFi-enabled devices using measurements of Received Signal Strength Indicator (RSSI) from WiFi probe requests. In this paper, which is an extension of the work initially presented in WiMob 2021, we continue the work on the mobility intelligence system (MobIntel) and study two broad approaches to tackle the problem of RSSI-based passive outdoor localization. One approach concerns multilateration and fingerprinting, both adapted from traditional active localization methods. For fingerprinting, we also show flaws in the previously reported area-under-the-curve method. The second approach uses machine learning, including machine learning-boosted multilateration, reference point classification, and coordinate regression. The localization performance of the two approaches is compared, and the machine learning methods consistently outperform the adapted traditional methods. This indicates that machine learning methods are promising tools for RSSI-based passive outdoor localization.

show abstract

Section: Fingerprintingmentioning

confidence: 99%