Artificial Intelligence-Based Energy Efficient Communication System for Intelligent Reflecting Surface-Driven VANETs
Qianqian Pan, Jun Wu, Jamel Nebhen, Ali Kashif Bashir, Yu Su, Jianhua Li
Abstract
The ever-increasing traffic, various delay-sensitive services, and energy consumption-constrained requirements have brought huge challenges to the current communication networks in the vehicular ad-hoc networks (VANETs). These challenges motivate academia and industry to investigate novel architectures with powerful data transmission and processing capabilities for low-latency and high energy-efficiency vehicular communication. In this paper, we propose an artificial intelligence (AI) and intelligent reflecting surface (IRS) empowered energy-efficiency communication system for VANETs. First, we design a smart and efficient hybrid vehicular communication framework, where IRS-aided dedicated short-range communication and long term evolution-based cellular communication are combined for data transmission in VANETs. Secondly, an IRS-aided data transmission is proposed to improve vehicular communication, in which the head vehicles selection method is designed. Based on the direct and IRS-reflecting signal propagation, fine-grained beamforming is achieved for directional vehicular transmission. Thirdly, a deep reinforcement learning (DRL) empowered network resource control and allocation scheme is proposed. In this scheme, we formulate an energy efficiency-maximizing model under the given transmission latency for VANETs and jointly optimize the settings of all participants to achieve efficient and low-latency communication. Finally, experimental results verify the effectiveness of our proposed communication system for VANETs.