Towards Practical Edge Security: Energy-Efficient RF Anomaly Detection on the NVIDIA Jetson Orin Nano

Download

ABSTRACT

Radio frequency (RF) communication has become
essential for critical infrastructure including healthcare, trans-
portation, and government services. While numerous ML models
have been developed for RF security, many require cloud com-
puting due to computational demands, introducing latency and
privacy concerns. With advances in AI edge computing, there is
growing opportunity to deploy models directly on edge devices.
This work presents the first benchmark study evaluating ML
models for RF anomaly detection on the NVIDIA Jetson Orin
Nano platform.

Nicholas D. Redmond is a senior at the University of Memphis pursuing a double major in Mathematics and Computer Science. Nicholas will begin his doctorate in Computer Science this fall of '26. Based in Memphis, Tennessee, his research focuses on RF signal security with applications extending to satellite and drone technology. Nicholas has presented his work at multiple academic venues, including two presentations at the Citadel, one at the University of Memphis, and is first author on the 2026 IEEE CCNC publication, "A Benchmark Study of RF Anomaly Detection Models on NVIDIA Jetson Orin Nano". His interdisciplinary approach combines creativity with practical applications in cybersecurity and wireless communications.
Submitted by Katie Dey on