How efficiently can AI recognize Wireless Devices?

Luangphairin, Waratchaya [Idaho National Laboratory]; Quach, Anna [Idaho National Laboratory]; Starks, Brandon [Idaho National Laboratory]

DOE OSTI · 3683988

How efficiently can AI recognize Wireless Devices?

Luangphairin, Waratchaya [Idaho National Laboratory]·Quach, Anna [Idaho National Laboratory]·Starks, Brandon [Idaho National Laboratory]

View original source ↗Read source document ↗Export metadata ↓

Abstract

This poster presents a hardware benchmarking methodology for a 3-layer CNN waveform classifier deployed using ONNX Runtime on an NVIDIA Jetson AGX Orin. The dataset consist of 9 signal types, -30 to +30 dB SNR with 5dB increments. Benchmarking on the Jetson AGX Orin gave an accuracy of 91.9% and GPU throughput of 107,120 predictions/sec (23× faster than CPU). The Jetson GPU reached approximately 27M samples/sec with stable performance but fell below the 40 MHz rate needed for real-time radio feeds. Sustained testing of 5 minutes confirmed stable performance with no memory leaks, establishing a reproducible benchmarking baseline for future edge-deployment optimization.

Explore related subjects

99 - GENERAL AND MISCELLANEOUS edge computing model benchmarking wireless signal classification

Keep this discovery

BibTeX RIS

Luangphairin, Waratchaya [Idaho National Laboratory], Quach, Anna [Idaho National Laboratory], Starks, Brandon [Idaho National Laboratory]. 2026-08-06. How efficiently can AI recognize Wireless Devices?. https://www.osti.gov/biblio/3683988

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

DOE OSTI2026

Predicting Atomistic Transitions with Transformers

Accurate knowledge of the atomistic transition pathways in materials and material surfaces is crucial for many material science problems. However, conventional simulation techniques used to find these transitions are extremely computationally intensive. Even with large-scale, accelerated material simulations, the computational cost constrains the applicable domain in practice. Machine learning models, with the potential to learn the complex emergent behaviors governing atomistic transitions as a fast surrogate model, have great promise to predict transitions with a vastly reduced computational cost. Here, we demonstrate how transformers can be trained to predict atomistic transitions in nano-clusters. We show how we evaluate physical validity of the predictions and how a multitude of additional, different microstates can be generated by slightly varying the data provided to the model.

36 MATERIALS SCIENCE↗

DOE OSTI2026

Offshore Gulf of Mexico Partnership for Carbon Storage--Resources and Technology Development GoMCarb

Final report of the GoMCarb Partnership's scientific and technical results.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2026

RFID-Enabled Tamper-Indicating Seal

Tamper-indicating seals protect sensitive materials, valuable equipment, and critical shipments. Most conventional seals must be checked through manual visual inspection and do not provide a digital record of their condition. Some attempts to remove, alter, or bypass a seal may also leave little visible damage, making tampering difficult to identify during routine inspections. Los Alamos National Laboratory has developed a smart tamper-indicating seal that converts even subtle tampering into a persistent digital record. The seal communicates its identity and status wirelessly through standard radio-frequency identification (RFID) readers, giving organizations a faster, more reliable way to verify that an asset has remained secure.

99 GENERAL AND MISCELLANEOUS↗