A Hybrid Quantum-Meta Reinforcement Learning and Graph Attention Transformer Approach for Low-Latency and a Secure D2D Routing in 6G Cellular Networks

This work involves the development of intelligent and security-oriented Device-to-Device (D2D) routing protocols specifically designed to handle dynamic, heterogeneous, and latency-critical communications in the emerging wireless communication environment, particularly for 6G setups. For instance, conventional routing solutions based on classical deep reinforcement learning and static optimization methodologies commonly fail to scale, adapt in real-time to dynamically changing topologies, and remain robust against security threats; thus, limiting their efficacy in the architectural frame of future 6G. However, this work focuses on introducing Quantum-Meta Learning Fusion with Graph Attention Transformer Networks for low-latency and secure D2D routing in 6G. The proposed framework is formed of three critical methodologies: (i) Quantum Meta Reinforcement Learning (QMRL), (ii) Graph Attention Transformer Network (GATN), and (iii) Evolutionary Graph Neural Network (EGNN). QMRL uses the parameterized quantum circuits to better balance the trade-off between exploration and exploitation. Additionally, a meta-learning module enables rapid adaptation to changing network conditions generated by gradient-based updates during end-to-end training. GATN combines graph attention mechanisms with transformer-based self-attention to capture spatial-temporal dependencies in network graphs; thus, the network will also be resilient to adversarial attacks. EGNN incorporates evolutionary approaches to enhance both node embeddings and routing policies, ensuring that traffic dynamics diversify over time and that link failures are appropriately addressed. Empirical analysis reveals that the proposed system will ultimately reduce latency by approximately 35-40%, improve the packet delivery ratio by 20%, and increase energy efficiency by 30-35% compared to classical models. Additionally, it increases resilience to jamming attacks by approximately 40%. Hence, these results substantiate the effectiveness of the fusion model in offering secure, adaptive, and low-latency routing solutions, making her a worthy candidate for future deployment in 6G D2D communications frameworks.