联合循环前缀和调制编码的链路自适应技术研究  

Research on Link Adaptation Technology of Joint Cyclic Prefix and MCS

作　　者：罗海鹏 刘玉军 刘潇然[1] 龚广伟[1] 赵海涛[1] LUO Hai-peng;LIU Yu-jun;LIU Xiao-ran;GONG Guang-wei;ZHAO Hai-tao(National University of Defense Technology,Changsha 410073,China;Army Academy of Armored Forces,Beijing 100072,China)

机构地区：[1]国防科技大学,湖南长沙410073 [2]陆军装甲兵学院,北京100072

出　　处：《中国电子科学研究院学报》2022年第7期657-663,共7页Journal of China Academy of Electronics and Information Technology

基　　金：国家自然科学基金重点项目(61931020);国家自然科学基金青年科学基金项目(62101569);国家自然科学基金联合基金项目(U19B2024);湖南省自然科学基金项目(2021JJ40690)。

摘　　要：为了应对未来无线通信使用场景日益多样化的趋势,需要进一步提升通信系统的自适应能力。在传统链路自适应技术基础上考虑波形设计维度,提出了一种联合正交频分复用(Orthogonal Frequency Division Multiplexing,OFDM)循环前缀与调制编码策略的链路自适应算法。所提算法根据当前信道条件,对接收后处理信噪比进行特征降维,采用多个二分类器结构预测不同系统参数的传输结果,选择使系统吞吐量最大的循环前缀和调制编码参数。同时,采用“探索/利用”的在线学习方式,适应信道的变化。仿真结果表明,所提算法能够适应时延扩展变化的信道环境,较传统链路自适应和随机策略取得近五倍吞吐量性能提升。此外,在链路质量突然恶化的情况下,所提在线算法能够在150 ms内重新达到较大的吞吐量性能。With the increasing diversity of usage scenarios of future wireless communications,it is necessary to further improve the adaptivity capability of the communication systems.This paper considers the link adaptation technology with waveform dimension.A new link adaptation strategy with joint optimizing the OFDM cyclic prefix and modulation and coding scheme is proposed.According to the channel condition,after feature dimension reduction of the receiver post processing SNRs,multiple binary-classifier structure is proposed to predict the transmission results of different parameters.Meanwhile,the exploration-exploitation approaches are used for online learning,which can adapt the changes of wireless channel.The simulation results show that the proposed algorithm can adapt to the channel environment with varying delay spread.It achieves nearly five times throughput performance improvement compared with the conventional method and random strategy.In addition,compared with the offline algorithm,the proposed online algorithm can adapt to the sudden deterioration of the link quality and return to a good throughput performance within 150 ms.

关 键 词：链路自适应 OFDM 机器学习 在线学习 

分 类 号：TN929[电子电信—通信与信息系统]