Year: 2025
Type: dissertation
Abstract: Wi-Fi has become integral to modern life, yet its pervasive connectivity introduces significant privacy risks as devices broadcast identifiable information during network discovery and communication.While MAC address randomization has been widely deployed to mitigate tracking, existing implementations remain vulnerable to sophisticated attacks that exploit temporal patterns, cross-layer features, and protocol-level information leakage.This dissertation presents a comprehensive investigation of privacy vulnerabilities in Wi-Fi MAC-layer protocols and proposes standards-compliant solutions to enhance user privacy.First, I demonstrate the Burst Interval Attack, which exploits predictable timing patterns in probe request transmissions to re-identify devices despite MAC randomization.To counter this, I introduce Jittery, a defense suite that anonymizes network discovery through bounded randomization of transmission timing while preserving connectivity and performance.Next, I present TRACE, a cross-layer fingerprinting attack that combines physical-layer signal characteristics with MAC-layer behavior to link randomized MAC addresses with high accuracy in real-world environments.Finally, I develop MIRAGE, a system enabling dynamic MAC randomization during active AP connections through multi-interface coordination, addressing a critical limitation where devices remain trackable throughout network sessions.Collectively, this research reveals that robust Wi-Fi privacy requires eliminating cross-layer linkability during network discovery and association.The proposed mechanisms demonstrate that standards-compliant solutions can achieve strong privacy guarantees without compromising connectivity or performance, contributing to a safer and more private wireless ecosystem.
Author: Cifuentes-Urtubey, Federico
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7220814151
