Researchers at Northeastern University’s Institute for Intelligent Networked Systems have developed GENESIS, an agentic AI framework that automates the entire R&D lifecycle of Radio Access Network (RAN) software for 5G/6G.
Here’s how it works: a researcher describes an idea in natural language, and GENESIS translates it into code, then independently tests, debugs, and optimizes it. This closes a long-standing engineering bottleneck: manually implementing a single 3GPP specification clause traditionally took months, while GENESIS accomplishes it in hours. The system feeds all results (code changes, test logs) back into a persistent knowledge base, improving over time.
Early successes include a program that improves phone-to-tower connectivity, reducing dropped calls. Future applications could enable satellite connectivity for smartphones in remote areas where terrestrial signals fail. Importantly, humans remain in the loop: safeguards allow researchers to monitor and halt the system if it deviates, mitigating concerns about autonomous AI going rogue. The team aims to commercialize GENESIS later this year.
About the Author
