LoRa Radio Frequency Fingerprinting Using a Hybrid Quantum-Classical Neural Network

To Truong An*, Simon L. Cotton, Junqing Zhang, Yuan Ding, Trung Q. Duong

*Corresponding author for this work

Research output: Contribution to conferencePaperpeer-review

4 Downloads (Pure)

Abstract

Radio frequency fingerprint identification is a promising technique for device authentication that relies on the unique radio frequency fingerprint features caused by hardware impairments. Existing radio frequency fingerprint identification models usually contain a significant number of trainable parameters, making them undesirable for Internet of Things applications. In this paper, we augment a classical neural network by introducing an intermediary quantum neural network stage to enhance the authentication of Internet of Things devices using radio frequency fingerprint features. The model is based on the combination of quantum and classical machine learning and benefits from a significantly smaller number of trainable parameters. Empirical results show that our proposed model not only achieves a much smaller footprint (in terms of device memory) but also delivers competitive accuracy to conventional deep learning approaches. It therefore shows much promise as a solution for securing networks which feature resource-constrained Internet of Things devices.
Original languageEnglish
Publication statusAccepted/In press - 8 Aug 2024
EventIEEE 100th Vehicular Technology Conference - Washington, United States
Duration: 7 Oct 202410 Oct 2024
https://events.vtsociety.org/vtc2024-fall/

Conference

ConferenceIEEE 100th Vehicular Technology Conference
Abbreviated titleVTC2024-Fall
Country/TerritoryUnited States
CityWashington
Period7/10/2410/10/24
Internet address

Fingerprint

Dive into the research topics of 'LoRa Radio Frequency Fingerprinting Using a Hybrid Quantum-Classical Neural Network'. Together they form a unique fingerprint.

Cite this