AI‑Driven Decoding of Graph‑Based Channel Codes via Reinforcement Learning (Systems and Methods for Decoding of Graph‑Based Channel Codes via Reinforcement Learning, Tech ID: 21‑021)
Technology Overview: This technology applies reinforcement learning (RL) to decode graph‑based channel codes used in modern digital communication systems. By learning optimal decoding strategies dynamically, the system improves error‑correction performance and adaptability compared to fixed‑algorithm decoders. This AI‑enabled approach is particularly valuable for next‑generation wireless and data communication systems operating under variable channel conditions.
Industry Pain Point: Conventional decoding algorithms struggle to adapt to dynamic channel environments.
NJIT Solution: Reinforcement learning enables adaptive, high‑performance decoding across varying conditions.
Key Features & Advantages
- AI‑driven adaptive decoding
- Improved error‑correction performance
- Applicable to modern channel codes
- Supports next‑generation communication systems
Development Stage: TRL 3–4 – Algorithmic validation completed.
Target Markets
- Wireless communications
- Data transmission systems
- Semiconductor and communications IP providers
Market Opportunity
- Global wireless communications market (2026): >$800B
- CAGR: ~6–7%
- Projected market size (2035): >$1.3T
Commercial & IP Details
Inventors: Allison Beemer, Salman Habib, Joerg Kliewer