WiFi Based Passive Vermin Sensing, Proof of Concept and Feasibility Analysis

Balanzategui, Benjamin (2023) WiFi Based Passive Vermin Sensing, Proof of Concept and Feasibility Analysis. [USQ Project]

[img] Text (Project - redacted)
Balanzategui_B_Leis_Redacted.pdf

Download (10MB)

Abstract

Vermin are a significant pest in Australia and throughout the world. When vermin ingress into a dwelling they are usually only detected after infestation has occurred. It is hypothesised that by analysing channel state information extracted from a WiFi network, it will be possible to detect and identify vermin when they are within the signal path of a typical WiFi network. A sensing utility function could be embedded into a WiFi system that detects vermin and alerts to their presence. This project determines if this concept is feasible by collecting channel state information from a WiFi system constructed using only commodity components.

Current standards defining the protocols that WiFi devices use require the determination of channel state information. Channel state information provides a rich representation of the propagation of individual components of the signal used in a WiFi system. Significant changes are observed when a physical object obstructs the signal path between transceivers. These changes can be analysed and categorised to identify the event occurring in the signal path, enabling passive sensing. Previous studies have investigated a variety of potential applications of WiFi sensing with an emphasis on health and wellbeing applications. The concept of using WiFi sensing to detect vermin is novel and has not previously been investigated.

To determine if WiFi based passive vermin sensing is feasible a WiFi network consisting of a single pair of transceivers was used to generate channel state information when a mouse is within the signal path. The collected channel state information was then analysed in comparison to channel state information collected from a control signal path, containing the same static objects but without a mouse present. The mouse caused conspicuous fluctuations in the magnitude measurements of the channel state information data and was able to be reliably identified by a Neural Network. The WiFi network utilised for testing was not modified in a way that inhibited normal communication functions. The findings of this project demonstrate that it is feasible to embed a sensing utility function into a typical WiFi network and vermin can be detected by a WiFi sensing system.


Statistics for USQ ePrint 52925
Statistics for this ePrint Item
Item Type: USQ Project
Item Status: Live Archive
Faculty/School / Institute/Centre: Current – Faculty of Health, Engineering and Sciences - School of Engineering (1 Jan 2022 -)
Supervisors: Leis, John
Qualification: Bachelor of Engineering
Date Deposited: 22 Sep 2025 04:54
Last Modified: 22 Sep 2025 04:55
URI: https://sear.unisq.edu.au/id/eprint/52925

Actions (login required)

View Item Archive Repository Staff Only