6G Deep Federated Optimization for Intelligent Data Quantum‐Resistant Routing and Monitoring System Using Game Theory

ABSTRACT The efficient path selection of data in wireless sensor networks is vital for enhancing total system performance. Traditional path selection protocols encounter challenges related to frequent node movements, optimization of energy efficiency, and the nature of network environments. To tackle these challenges, an innovative methodology termed Deep reinforcement arctic puffin federated stochastic gradient learning (DRAFSGL) integrates federated learning and puffin optimization for energy‐efficient multipath routing. Multiview deep embedded clustering (MDEC), enhanced by coyote and badger optimization into efficient clusters to reduce routing complexity. To ensure privacy, quantum‐resistant homomorphic encryption (QRHE) enables secure computation without safeguarding against quantum threats. The data monitoring system applies game theory to optimize the behavior of agents involved in monitoring activities in the system. The suggested approach has exhibited extraordinary evaluation, attaining a peak accuracy of 99.7% alongside a precision of 98.9%, a specificity of 98.4%, a delay of 18 ms, a throughput of 87.9 Kbps, and a remarkable recall rate of 98.9%, as well as an F1‐score when compared to existing methods. Overall, DRAFSGL methodology improves adaptive, secure, and energy‐efficient multipath routing through the combination of federated learning and arctic puffin optimization. Meanwhile, the MDEC technique with improved coyote and badger optimization (ICBO) simplifies routing complexity and QRHE guarantees quantum‐resistant, secure data transmission, effectively addressing the shortcomings of current techniques in dynamic 6G wireless sensor network (WSN)‐based Internet of Things (IoT) systems.