Author(s)
Nancy Joseph, Shonda Bernadin, Hongmei Chi, andTejal Mulay
Abstract
Large language models (LLMs) have become widely available, resulting in a global phenomenon that has promoted their use across all spheres of society. Researchers are attempting to utilize LLMs for spectrum management to reduce workload due to their ability to process vast amounts of data. As the world transitions into 5G and eventually 6G, spectrum management will need a system to handle large volumes of heterogeneous data sources—ranging from technical standards and regulatory documents to real-time performance metrics. LLMs integrated with Retrieval-Augmented Generation (RAG) frameworks present a novel opportunity to enable context-aware and adaptive spectrum management strategies. This project aims to evaluate the different LLM-RAG models on policy documents useful for spectrum management.