Meta-Reinforced Graph-Aware Secure SDN Framework Using Hybrid GWCC-Chaos Cryptography and Energy-Aware Routing

The rapid growth of Internet of Things (IoT) devices, 5G connectivity, and data-driven services has increased the demand for intelligent, secure, and energy-efficient network infrastructures. Static routing choices, inflexible behavior, and high processing requirements of security schemes pose challenges in traditional Software-Defined Networking (SDN) models. In this paper, we present a new SDN framework in which Meta-Reinforcement Learning (Meta-RL) and Graph Attention Networks (GATs) are combined to support dynamic, topology-aware, and energy-efficient routing. Meta-RL enables the SDN controller to quickly adapt to changes in network conditions through prior knowledge, and GAT improves the learning process by focusing on the most relevant topological characteristics. The framework employs a hybrid cryptographic model based on Genus Weierstrass Curve Cryptography (GWCC), combined with chaotic map encryption to provide secure data transmission without adversely affecting real-time performance. The combined architecture enhances entropy and reduces processing latency. The proposed system is implemented and tested on a Mininet-based platform with a multi-hop structure using an OpenFlow topology. The results show significant improvements in throughput, latency, packet delivery ratio, energy efficiency, and Area Under the Curve (AUC) compared with current models, including Deep Q-Network (DQN), Q-learning-based Routing (QLR), and Elliptic Curve Cryptography (ECC)-based routing.