Advancing communication networks: integrating enhanced knowledge mapping with hybrid deep recurrent neural networks for dynamic spectrum access
Sravana Kumar Yelithoti, Tapaswini Samant, Swati Swayamsiddha · International Journal of Basic and Applied Sciences · 2025
By merging knowledge mapping using Hybrid Deep Recurrent Neural Networks (RNNs), our suggested method maximizes dynamic spectrum access in diverse networks. Optimizing the assignment of spectrum resources while enabling different device characteristics and network circumstances is our approach to addressing the issues of spectrum allocation in different network situations. Improved processing capabilities at the network's edge enable real time monitoring of patterns in spectrum consumption. In order to dynamically allocate spectrum according to user demands and network dynamics, our dynamic spectrum access technology employs reinforcement learning algorithms. Hybrid Deep RNNs take use of both deep learning as well as recurrent neural networks to enhance feature extraction and behavioral dependence modeling in spectrum data. In order to guarantee the system's reliability and resilience in real world applications, assessment indicators are used to analyze its performance and efficiency. Consistent with our hypothesis, the results demonstrate significant gains in spectrum utilization effectiveness and allocation accuracy, validating our approach to maximizing resource consumption and facilitating faultless functioning in diverse network settings.