5G-Based Mobile Communications: Stable Route Selection for Adaptive Packet Transmission

Abstract

A major difficulty in 5G-based mobile communications is to guarantee robust selecting routes for adaptive packet transmission in a dynamic environment. Traditional routing protocols struggle to adapt to the fluctuating wireless channel conditions inherent in 5G networks. To address this, our study introduces a novel system that integrates Deep Q-Networks (DQN) techniques with the Zone routing protocol (ZRP). Leveraging real-time network data including channel quality, traffic load, and congestion levels, the system employs machine learning algorithms to predict route stability. This predictive capability enables dynamic identification of the most stable route for packet transmission, with continuous monitoring and adjustment in response to evolving network conditions. Our proposed system follows a multi-step flow, starting from data collection and culminating in route selection based on machine learning predictions. Extensive simulations and real-world experiments validate the efficacy of our approach, demonstrating significant improvements in packet delivery ratio, latency, and overall network stability compared to conventional methods. Notably, our system exhibits resilience against varying network conditions and maintains scalability with increasing network size and traffic load. Through the fusion of machine learning and routing protocols, our study offers a promising solution to the critical challenge of stable route selection in 5G-based mobile communications, addressing the diverse demands of emerging applications and services. © 2024, Scibulcom Ltd.. All rights reserved.