基于LSTM的多站点协作波束切换算法  

LSTM-based Multi-site Collaborative Beam Switching Algorithm

作　　者：王海洋[1] 何兴文 孙海蓬[1] 王其盛 苏俊浩 张志龙 WANG Hai-yang;HE Xing-wen;SUN Hai-peng;WANG Qi-sheng;SU Jun-hao;ZHANG Zhi-long(Shandong Electric Power Engineering Consulting Institute Co.,Ltd.,Jinan 250100,China;ZTE Corporation,Shanghai 200120,China;School of Information and Communication Engineering,Beijing University of Posts and Telecommunications,Beijing 100876,China)

机构地区：[1]山东电力工程咨询院有限公司,山东济南250100 [2]中兴通讯股份有限公司,上海200120 [3]北京邮电大学信息与通信工程学院,北京100876

出　　处：《中国电子科学研究院学报》2024年第3期239-247,共9页Journal of China Academy of Electronics and Information Technology

摘　　要：毫米波频段拥有非常丰富的频谱资源,支持更高速的数据传输,因此成为通信领域近年来的研究热点。基于大规模输入输出(Massive Multiple Input Multiple Output,mMIMO)的波束成形技术可以有效补偿毫米波的高路损,是5G毫米波通信中的一项关键技术。但是在高速移动场景下,信道变化加速,必须进行频繁的波束切换才能保证良好的通信质量,故而大大增加了信令开销与资源消耗。如何在高动态场景下降低波束切换开销、提升波束切换速度与效率成为关键问题。针对直射径(Light of Sight,LOS)被遮挡情况下的复杂信道环境,文中提出了一种基于长短期记忆(Long Short Term Memory,LSTM)网络的多站点协作波束切换算法,根据过去一段时间的波束信息,学习信道环境与服务波束序号的内在联系,并预测当前时刻切换的最佳波束。仿真结果表明,与基于当前信道环境信息的遍历搜索算法相比,所提波束切换算法在上行发射功率较高时性能接近,在上行发送功率较低时,性能更优,体现了所提算法的优势。The millimeter wave frequency band boasts abundant spectrum resources,supporting faster data transmission,making it a recent focal point in the field of communication.Beamforming based on Massive Multiple Input Multiple Output(mMIMO) is a crucial technology in 5G millimeter wave communication,since it effectively compensates for the high path loss associated with millimeter waves.However,in high-speed mobile scenarios,channel conditions change rapidly,necessitating frequent beam switching to ensure good communication quality.This significantly increases signaling overhead and resource consumption.Thus,the key challenge is to reduce beam switching overhead and improve the speed and efficiency of beam switching in high-dynamic scenarios.In this paper,addressing complex channel environments when line-of-sight(LOS) is obstructed,we propose a multi-site collaborative beam switching algorithm based on Long Short Term Memory(LSTM) networks.It learns the inherent relationship between channel conditions and the sequence of service beams based on past beam information and predicts a beam to switch to at the current time.Simulation results indicate that,compared to the exhaustive search algorithm based on the current channel environment information,our proposed beam switching algorithm performs similarly when the uplink transmission power is high and even better when the uplink transmission power is low.This demonstrates the advantages of the proposed algorithm.

关 键 词：波束切换 站点协作 长短期记忆网络 

分 类 号：TN929.5[电子电信—通信与信息系统]