From April 29 to May 2, Li Guyue, lecturer of School of Cyber Science and Engineering and Yu Jiabao, doctoral candidate enrolled in 2014 were invited to INFOCOM2019 Workshop (IEEE International Conference on Computer Communications 2019: The 1st International Workshop on Internet of Things for Adversarial Environments) hosted in Paris, France. At this conference, they shared their research findings on physical layer security laboratory (PLSLAB), A Multi-Sampling Convolutional Neural Network-Based RF Fingerprinting Approach for Low-Power Devices, at the session of Security, Robustness, and Load Management and had in-depth exchanges with scholars from all around the world. The corresponding author of the paper is Hu Aiqun, professor of School of Cyber Science and Engineering, the first author is Yu Jiabao, and the lead authors are lecturer Li Guyue and associate professor Peng Linning.
This paper studies the impact of sleep mode switching of low-power devices on existing radio frequency (RF) fingerprinting techniques and presents an adaptive region of interest (ROI) selection strategy to deal with the semi-steady behavior due to sleep mode switching. Moreover, based on the current single-sampling RF fingerprinting approach to feature extraction, this paper provides a multiple downsampling convolutional neural network to extract RF fingerprint and extensive experiments with 54 ZigBee devices as targets are conducted with the identification accuracy as high as 97%.
A Multi-Sampling Convolutional Neural Network-Based RF Fingerprinting Approach for Low-Power Devices
Authors: Jiabao Yu, Aiqun Hu, Guyue Li and Linning Peng (Southeast University, P.R. China)
The link to the paper is as follows:
The expansion of the conference paper has been published in IEEE Internet of Things Journal (JCR-1, IF=5.863). The link to the paper is as follows: https://ieeexplore.ieee.org/document/8692418
INFOCOM (IEEE International Conference on Computer Communications) is a flagship conference organized by IEEE on communications network as well as a landmark conference in the international field of communications network at present. Hosted by IEEE Communications Society, INFOCOM, together with MOBICOM and SIGCOMM, is listed as the type-A conference on computer networks recommended by China Computer Federation (CCF). This highly influential conference provides a great show stage and exchange platform for scholars in network communications worldwide. Having set a high requirement on the quality of papers over the years, INFOCOM boasts a strong judge team that consists of nearly all first-class researchers on network communications across the world. INFOCOM is not only a platform to release research findings, but also a grand gathering for researchers in this field to know each other, make friends, and have a discussion.