People density estimation using Wi-Fi infrastructure
Author
Buchakchiev, Radoslav Naskov
Term
4. term
Education
Publication year
2016
Submitted on
2016-06-02
Abstract
This thesis explores whether people counts can be estimated by leveraging existing Wi‑Fi infrastructure rather than deploying dedicated, costly sensors. It poses and addresses the question: Can passive collection of 802.11 probe requests from smartphones be used to assess people density—specifically bus passenger counts—in a reliable, low‑cost manner? A prototype is designed and implemented, comprising a scanner that captures Wi‑Fi probe requests and an algorithm that filters and aggregates the captured signals to estimate the number of people. The algorithm is further enhanced by incorporating received signal strength (RSSI). The system is tested on a city bus, and results indicate it can estimate the number of people with some limitations in accuracy. Considering its low cost, the approach is promising, and several directions are suggested for future work to improve estimation accuracy.
Denne afhandling undersøger, om antal personer kan estimeres ved at udnytte eksisterende Wi‑Fi‑infrastruktur i stedet for at installere dyrt udstyr. Projektet formulerer og adresserer spørgsmålet: Kan passiv indsamling af 802.11 probeanmodninger fra smartphones bruges til at vurdere tæthed af mennesker, konkret passagerer i en bybus, på en pålidelig og lavpris måde? Der udvikles og implementeres en prototype bestående af en scanner, der registrerer Wi‑Fi probeanmodninger, samt en algoritme, der filtrerer og sammenfatter de registrerede signaler for at estimere antallet af personer. Algoritmen forbedres yderligere ved at inddrage modtaget signalstyrke (RSSI). Systemet afprøves i en virkelig bybus, og resultaterne viser, at antallet af personer kan estimeres med visse begrænsninger i nøjagtigheden. Set i lyset af den lave omkostning fremstår løsningen som lovende, og der peges på flere retninger for fremtidigt arbejde, der kan forbedre nøjagtigheden.
[This apstract has been generated with the help of AI directly from the project full text]
Keywords
Wi-Fi ; probe request ; infrastructure ; bus