Transformer-based radio signal classifier for spectrum
This technology uses two cooperating neural networks to learn how to identify and characterize radio signals more accurately. A 'signal transformer' sits between the two networks: it takes a raw radio signal plus the first network's initial analysis, then applies mathematical transforms to create an improved version of the signal before feeding it to the second network. The two networks train together, with errors from the final output used to improve both simultaneously. The result is a machine learning system that gets better at recognizing signal types, modulations, or other characteristics even in noisy or complex radio environments.
What you could build
An embedded or cloud-based radio signal intelligence module for spectrum monitoring, interference detection, or automatic modulation classification — sold to defense contractors, spectrum regulators, or wireless equipment makers who need to identify signals without prior knowledge of their source.
Who in Virginia should care
Northern Virginia defense contractors and signals intelligence firms (e.g., SAIC, Leidos, Booz Allen) working on electronic warfare or spectrum management would have direct interest.
Readiness: Lab validated
Concept — described but not yet demonstrated. Lab validated — supported by experimental results in the patent. Prototype likely — the text describes a built, working embodiment.
Readiness is inferred from the patent text, not from a lab visit.
The record
- Inventors
- Timothy James O`Shea
- Granted
- May 21, 2019
- Status
- Granted patent
- Patent number
- 10296831
Ready to talk?
Virginia Tech Intellectual Properties handles licensing for this technology.
Prosim summaries are generated from public patent text and are not legal advice.