Energy-efficient secure software-defined networking with reinforcement learning and Weierstrass cryptography

Abstract

In the age of rapidly advancing 5G connectivity, artificial intelligence (AI), and the internet of things (IoT), network data has grown enormously, demanding more efficient and secure management solutions. Traditional networking systems, limited by manual controls and static environments, are unable to fulfill the dynamic demands of modern internet services. This paper proposes an innovative software-defined networking (SDN) framework that utilizes exponential spline regression reinforcement learning (ESR-RL) with genus Weierstrass curve cryptography (GWCC) to boost energy efficiency and data security. The ESR-RL algorithm reliably anticipates network traffic patterns, optimizing path selection to enhance routing efficiency while minimizing consumption of energy. GWCC also enables strong encryption and decryption, considerably increasing data security without impacting system performance. To further improve network reliability, the Skellam distributed Siberian TIGER optimization algorithm (SDSTOA) is used to dynamically acquire features and balance loads, resulting in optimal network performance. Extensive simulations show that the proposed framework performs better than existing models in terms of accuracy, precision, recall, F-measure, sensitivity, and specificity. Improvements in latency, turnaround time, and network throughput demonstrate the framework's success. This scalable and adaptive technology establishes a new standard for SDN systems by providing a safe, energy-efficient, and performance-optimized strategy for future network infrastructures.