Scholarlink

<p>Vehicle-to-Everything (V2X) connectivity has emerged as a key enabler in Fifth Generation (5G) and Sixth Generation (6G) mobile networks, aiming to address critical challenges in transportation systems while enhancing road safety, traffic efficiency, and user comfort. However, the performance of V2X communication systems, particularly for safety-critical applications, remains constrained by limited spectral resources, massive device connectivity, unstable wireless links, and high signaling overhead associated with centralized resource management. To address these challenges, this paper proposes a cellular Non-Orthogonal Multiple Access (NOMA)-based resource allocation framework leveraging Deep Reinforcement Learning (DRL). Specifically, we develop a Proximal Policy Optimization (PPO)-based algorithm that dynamically allocates radio Resource Blocks (RBs) while accounting for the heterogeneous requirements of safety and non-safety V2X traffic. Unlike existing approaches, the proposed method jointly considers traffic prioritization, spectrum efficiency, and learning-based decision-making within a unified framework, enabling efficient and scalable resource allocation under dynamic vehicular conditions. Simulation results demonstrate that the proposed PPO-based scheme consistently outperforms conventional DRL and state-of-the-art baseline algorithms in terms of average throughput, Quality of Service (QoS), and Average Blocking Rate (ABR), highlighting its effectiveness in supporting reliable and efficient V2X communications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

NOMA Resource Allocation Algorithm Based on Deep Reinforcement Learning for V2X Networks

  • 期刊论文

  • 发表日期：

    2026年5月8日

查看全文
  • Ala Din Trabelsi,
  • Arwa Amaira,
  • Faouzi Zarai

摘要

Vehicle-to-Everything (V2X) connectivity has emerged as a key enabler in Fifth Generation (5G) and Sixth Generation (6G) mobile networks, aiming to address critical challenges in transportation systems while enhancing road safety, traffic efficiency, and user comfort. However, the performance of V2X communication systems, particularly for safety-critical applications, remains constrained by limited spectral resources, massive device connectivity, unstable wireless links, and high signaling overhead associated with centralized resource management. To address these challenges, this paper proposes a cellular Non-Orthogonal Multiple Access (NOMA)-based resource allocation framework leveraging Deep Reinforcement Learning (DRL). Specifically, we develop a Proximal Policy Optimization (PPO)-based algorithm that dynamically allocates radio Resource Blocks (RBs) while accounting for the heterogeneous requirements of safety and non-safety V2X traffic. Unlike existing approaches, the proposed method jointly considers traffic prioritization, spectrum efficiency, and learning-based decision-making within a unified framework, enabling efficient and scalable resource allocation under dynamic vehicular conditions. Simulation results demonstrate that the proposed PPO-based scheme consistently outperforms conventional DRL and state-of-the-art baseline algorithms in terms of average throughput, Quality of Service (QoS), and Average Blocking Rate (ABR), highlighting its effectiveness in supporting reliable and efficient V2X communications.