DOI

https://doi.org/10.25772/SGZX-PE20

Author ORCID Identifier

0000-0001-6386-5704

Defense Date

2023

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Computer Science

First Advisor

Eyuphan Bulut

Abstract

WiFi sensing offers a powerful method for tracking physical activities using the radio-frequency signals already found throughout our homes and offices. This novel sensing modality offers continuous and non-intrusive activity tracking since sensing can be performed (i) without requiring wearable sensors, (ii) outside the line-of-sight, and even (iii) through the wall. Furthermore, WiFi has become a ubiquitous technology in our computers, our smartphones, and even in low-cost Internet of Things devices. In this work, we consider how the ubiquity of these low-cost WiFi devices offer an unparalleled opportunity for improving the scalability of wireless sensing systems. Thus far, WiFi sensing research assumes costly offline computing resources and hardware for training machine learning models and for performing model inference. To improve the scalability of WiFi sensing systems, this dissertation introduces techniques for improving machine learning at the edge by thoroughly surveying and evaluating signal preprocessing and edge machine learning techniques. Additionally, we introduce the use of federated learning for collaboratively training machine learning models with WiFi data only available on edge devices. We then consider privacy and security concerns of WiFi sensing by demonstrating possible adversarial surveillance attacks. To combat these attacks, we propose a method for leveraging spatially distributed antennas to prevent eavesdroppers from performing adversarial surveillance while still enabling and even improving the sensing capabilities of allowed WiFi sensing devices within our environments. The overall goal throughout this work is to demonstrate that WiFi sensing can become a ubiquitous and secure sensing option through the use of on-device computation on low-cost edge devices.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

5-9-2023

Download

Share

COinS