Optimizing Spectrum Access in Cognitive Radio Networks: A Reinforcement Learning Approach with Correlation-Aware Utility Maximization

The increasing demand for wireless communication across commercial and military domains has created a critical need for efficient spectrum usage. This paper proposes a novel reinforcement learning-based approach to optimize spectrum access in Cognitive Radio Networks (CRNs). Unlike existing models, our method integrates spatial and temporal correlation through a partially observable Markov decision process (POMDP) and leverages the PERSEUS algorithm for efficient value iteration. Simulation results demonstrate improved secondary user throughput and reduced primary user interference compared to traditional methods such as MAP estimators and correlation-based clustering. The model's scalability and practical applicability make it a promising solution for next-generation dynamic spectrum access.