摘要
随着5G无线通信技术和物联网技术的发展,海量通信终端设备连接及受限的频谱资源等问题给高速率大容量无线通信带来了严峻挑战。智能反射面通信技术作为一种新兴的无线通信技术以其无源、低成本的特点吸引了人们广泛的关注。提出了一种新的优化算法,将深度学习技术与智能反射面相结合,通过训练神经网络建立信道状态信息与智能反射面的最优反射系数矩阵之间的映射关系,在保护物联网数据隐私的同时,实现智能反射面的实时重配置进而提升接收端的用户通信速率。
With the development of 5th-generation(5G) wireless systems and internet of things(IoT), high speed wireless communications are facing severe challenges due to numerous connections of communication terminals and limited frequency resources. Intelligent reflecting surface(IRS), a newly appeared wireless communication technology, has attracts people′s attentions worldwide by its low power consumptions and costs. In order to improve the communication rates of users, this paper proposes a novel optimization algorithm by implementing the deep learning to IRS for establishing the mapping from the channel state information to the optimal reflecting coefficient matrix of IRS. The present algorithm can perform real-time reconfiguration of IRS while protecting the privacy of IoT users.
作者
李苗钰
杜忠昊
刘雨彤
牛思莹
LI Miaoyu;DU Zhonghao;LIU Yutong;NIU Siying(School of Information Science and Technology, Northwestern University, Xi′an 710127, China)
出处
《西北工业大学学报》
EI
CAS
CSCD
北大核心
2021年第2期454-461,共8页
Journal of Northwestern Polytechnical University
基金
国家自然科学基金(62002291)资助。
关键词
智能反射面
物联网
深度学习
联邦学习
服务质量保障
intelligent reflecting surface
internet of things
deep learning
federated learning
privacy protection