AIAS: Experimental Evaluation of Spectrum Utilization Over Cognitive Radio Networks Using Artificial Intelligence (AI) Assisted Adaptive Searching Algorithm

The AI-Assisted Adaptive Searching Algorithm (AIAS) enhances spectrum utilization in Cognitive Radio Networks (CRNs) by predicting and dynamically allocating available frequency bands to secondary users. Using the Kaggle Spectrum Dataset, AIAS employs supervised learning to forecast spectrum availability with a 91.3% average accuracy, outperforming traditional searching methods by approximately 13%. Additionally, reinforcement learning enables real-time decision-making for optimal frequency allocation, achieving an average spectrum utilization efficiency of 92.5% and interference-free allocations up to 95.1%. Compared to conventional algorithms, AIAS reduces latency by an average of 50%, achieving allocation times of approximately 15.2 milliseconds. These improvements demonstrate AIAS's capability to adapt in spectrum-dense environments, ensuring minimal interference with primary users and higher spectrum utilization efficiency. This study highlights the potential of AIAS to address the limitations of static allocation in CRNs and underscores the advantages of machine learning in dynamic spectrum management. Future research could focus on integrating advanced AI models, such as deep learning, to further optimize prediction accuracy and enhance the robustness of adaptive spectrum utilization in increasingly complex network environments.