当前位置: X-MOL 学术Veh. Commun. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Priority based V2V data offloading scheme for FiWi based vehicular network using reinforcement learning
Vehicular Communications ( IF 5.8 ) Pub Date : 2023-06-09 , DOI: 10.1016/j.vehcom.2023.100629
Akshita Gupta , Saurabh Jaiswal , Vivek Ashok Bohara , Anand Srivastava

Intelligent vehicular network is one of the popular paradigms for sixth-generation (6G) networks. With the growth in vehicular traffic density, it is essential to ensure safety and security by providing ultra reliable low latency (URLL) services for vehicular communication. Moreover, along with the URLL services, it is also essential that these services are energy-efficient. These requirements can be fulfilled by the integration of optical fiber at the back-end with wireless front-end, commonly referred to as fiber-wireless (FiWi) networks. In this work, we consider a FiWi-based vehicular network that consists of a optical fiber back-end network with wireless vehicular-to-infrastructure (V2I) network at the front-end. Specifically, to ensure that high priority vehicular traffic has lower latency and energy consumption, we propose a reinforcement learning (RL) based double deep Q-network (DDQN) scheme that chooses the cluster-head for vehicular communication. The proposed algorithm uses energy and latency aware cluster-head selection that meets the intelligent transportation system (ITS) requirements. The proposed algorithm is compared with the non-RL-based approach such as, random and max-queue based cluster-head selection. The simulation results demonstrate the efficiency of the proposed algorithm in terms of energy, latency, throughput, and reliability.



中文翻译:

使用强化学习的基于 FiWi 的车辆网络的基于优先级的 V2V 数据卸载方案

智能车载网络是第六代(6G)网络的流行范例之一。随着车辆交通密度的增长,通过为车辆通信提供超可靠的低延迟(URLL)服务来确保安全至关重要。此外,与 URLL 服务一样,这些服务的节能也至关重要。这些要求可以通过光纤集成来满足后端带有无线前端,通常称为光纤无线(FiWi)网络。在这项工作中,我们考虑一个基于 FiWi 的车载网络,该网络由光纤后端网络和前端无线车辆到基础设施 (V2I) 网络组成。具体来说,为了确保高优先级车辆交通具有较低的延迟和能耗,我们提出了一种基于强化学习(RL)的双深度Q-网络(DDQN)方案,选择用于车辆通信的簇头。所提出的算法使用满足智能交通系统(ITS)要求的能量和延迟感知簇头选择。将所提出的算法与非基于强化学习的方法(例如随机和基于最大队列的簇头选择)进行比较。仿真结果证明了该算法在能量、延迟、吞吐量和可靠性方面的效率。

更新日期:2023-06-09
down
wechat
bug