Author(s): Jitendra Agrawa¹, Virendra Kumar Tiwari², Sanjay Thakur³

Abstract:The advent of 6G networks brings unprecedented demands for ultra-low latency, massive connectivity, and intelligent, adaptive services, posing significant challenges for traditional traffic management methods. This paper presents a novel AI-driven traffic engineering framework integrated within Software-Defined Networking (SDN) architecture to meet these demands. By leveraging Deep Reinforcement Learning (DRL), the proposed system autonomously learns optimal routing strategies, dynamically allocates bandwidth, and effectively mitigates network congestion under diverse conditions. Simulations conducted using the Mininet emulator and TensorFlow-based DRL models demonstrate up to 35% reduction in latency and a 42% improvement in bandwidth utilization compared to conventional traffic engineering approaches. The results highlight the potential of combining SDN with AI to enable intelligent, real-time traffic optimization for future ultra-dense 6G networks.

Keywords: 6G Networks, Traffic Engineering, Software-Defined Networking (SDN), Deep Reinforcement Learning (DRL), Deep Q-Network (DQN), Quality of Service (QoS), Adaptive Routing, Latency Optimization.

DOI: 10.61165/sk.publisher.v12i7.3



Download Full Article from below:

AI-Driven Traffic Engineering for 6G Networks: A Deep Reinforcement Learning Approach in SDN Architecture

Pages:21-38