AI-learned encoder and channel estimator for MIMO wireless
This technology uses machine learning to make wireless communication systems smarter about how they send and receive signals over MIMO channels — the multi-antenna setups that underpin modern 4G, 5G, and Wi-Fi networks. The key innovation is training neural networks to handle two hard problems simultaneously: encoding information for transmission and decoding it on arrival, while also estimating the current state of the wireless channel (how signals are being distorted, reflected, or attenuated in real time). The system learns end-to-end, meaning the transmitter and receiver are jointly optimized rather than engineered as separate fixed components. This allows the system to adapt to complex, real-world channel conditions that traditional signal processing rules struggle to handle efficiently.
What you could build
A software or firmware module for 5G/6G base stations and user devices that replaces or augments conventional signal processing stacks with learned encoders and channel estimators, improving throughput and reliability; buyers would be wireless chipset makers, telecom equipment vendors, and network operators.
Who in Virginia should care
Northern Virginia's dense concentration of defense contractors, telecom infrastructure operators, and federal agencies with spectrum interests (DARPA, DoD, FCC proximity) makes this directly relevant.
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, Tugba Erpek
- Granted
- June 17, 2025
- Status
- Granted patent
- Patent number
- 12334997
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.