Distinguishable IQ Feature Representation for Domain-Adaptation Learning of WiFi Device Fingerprints
Deep learning (DL)-based RF fingerprinting (RFFP) technology has emerged as a powerful physical-layer security mechanism, enabling device identification and authentication based on unique device-specific signatures that can be extracted from the received RF signals. However, DL-based RFFP methods fa...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Transactions on Machine Learning in Communications and Networking |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10640139/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850227224099160064 |
|---|---|
| author | Abdurrahman Elmaghbub Bechir Hamdaoui |
| author_facet | Abdurrahman Elmaghbub Bechir Hamdaoui |
| author_sort | Abdurrahman Elmaghbub |
| collection | DOAJ |
| description | Deep learning (DL)-based RF fingerprinting (RFFP) technology has emerged as a powerful physical-layer security mechanism, enabling device identification and authentication based on unique device-specific signatures that can be extracted from the received RF signals. However, DL-based RFFP methods face major challenges concerning their ability to adapt to domain (e.g., day/time, location, channel, etc.) changes and variability. This work proposes a novel IQ data representation and feature design, termed Double-Sided Envelope Power Spectrum or <monospace>EPS</monospace>, that is proven to significantly overcome the domain adaptation challenges associated with WiFi transmitter fingerprinting. By accurately capturing device hardware impairments while suppressing irrelevant domain information, <monospace>EPS</monospace> offers improved feature selection for DL models in RFFP. Our experimental evaluation demonstrates the effectiveness of the integration of <monospace>EPS</monospace> representation with a Convolution Neural Network (CNN) model, termed <monospace>EPS-CNN</monospace>, achieving over 99% testing accuracy in same-day/channel/location evaluations and 93% accuracy in cross-day evaluations, outperforming the traditional IQ representation. Additionally, <monospace>EPS-CNN</monospace> excels in cross-location evaluations, achieving a 95% accuracy. The proposed representation significantly enhances the robustness and generalizability of DL-based RFFP methods, thereby presenting a transformative solution to IQ data-based device fingerprinting. |
| format | Article |
| id | doaj-art-4a58c0f0413247bfb57ef7ed71756865 |
| institution | OA Journals |
| issn | 2831-316X |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Transactions on Machine Learning in Communications and Networking |
| spelling | doaj-art-4a58c0f0413247bfb57ef7ed717568652025-08-20T02:04:54ZengIEEEIEEE Transactions on Machine Learning in Communications and Networking2831-316X2024-01-0121404142310.1109/TMLCN.2024.344674310640139Distinguishable IQ Feature Representation for Domain-Adaptation Learning of WiFi Device FingerprintsAbdurrahman Elmaghbub0https://orcid.org/0000-0003-3704-6056Bechir Hamdaoui1https://orcid.org/0000-0002-6085-4505Oregon State University, Corvallis, OR, USAOregon State University, Corvallis, OR, USADeep learning (DL)-based RF fingerprinting (RFFP) technology has emerged as a powerful physical-layer security mechanism, enabling device identification and authentication based on unique device-specific signatures that can be extracted from the received RF signals. However, DL-based RFFP methods face major challenges concerning their ability to adapt to domain (e.g., day/time, location, channel, etc.) changes and variability. This work proposes a novel IQ data representation and feature design, termed Double-Sided Envelope Power Spectrum or <monospace>EPS</monospace>, that is proven to significantly overcome the domain adaptation challenges associated with WiFi transmitter fingerprinting. By accurately capturing device hardware impairments while suppressing irrelevant domain information, <monospace>EPS</monospace> offers improved feature selection for DL models in RFFP. Our experimental evaluation demonstrates the effectiveness of the integration of <monospace>EPS</monospace> representation with a Convolution Neural Network (CNN) model, termed <monospace>EPS-CNN</monospace>, achieving over 99% testing accuracy in same-day/channel/location evaluations and 93% accuracy in cross-day evaluations, outperforming the traditional IQ representation. Additionally, <monospace>EPS-CNN</monospace> excels in cross-location evaluations, achieving a 95% accuracy. The proposed representation significantly enhances the robustness and generalizability of DL-based RFFP methods, thereby presenting a transformative solution to IQ data-based device fingerprinting.https://ieeexplore.ieee.org/document/10640139/RF/device fingerprintingdomain adaptationRF datasetsdeep learning feature designoscillatorsRF data representation |
| spellingShingle | Abdurrahman Elmaghbub Bechir Hamdaoui Distinguishable IQ Feature Representation for Domain-Adaptation Learning of WiFi Device Fingerprints IEEE Transactions on Machine Learning in Communications and Networking RF/device fingerprinting domain adaptation RF datasets deep learning feature design oscillators RF data representation |
| title | Distinguishable IQ Feature Representation for Domain-Adaptation Learning of WiFi Device Fingerprints |
| title_full | Distinguishable IQ Feature Representation for Domain-Adaptation Learning of WiFi Device Fingerprints |
| title_fullStr | Distinguishable IQ Feature Representation for Domain-Adaptation Learning of WiFi Device Fingerprints |
| title_full_unstemmed | Distinguishable IQ Feature Representation for Domain-Adaptation Learning of WiFi Device Fingerprints |
| title_short | Distinguishable IQ Feature Representation for Domain-Adaptation Learning of WiFi Device Fingerprints |
| title_sort | distinguishable iq feature representation for domain adaptation learning of wifi device fingerprints |
| topic | RF/device fingerprinting domain adaptation RF datasets deep learning feature design oscillators RF data representation |
| url | https://ieeexplore.ieee.org/document/10640139/ |
| work_keys_str_mv | AT abdurrahmanelmaghbub distinguishableiqfeaturerepresentationfordomainadaptationlearningofwifidevicefingerprints AT bechirhamdaoui distinguishableiqfeaturerepresentationfordomainadaptationlearningofwifidevicefingerprints |