高效的基于数据与模型的信道估计算法  

Efficient model-and-data based channel estimation algorithm

作　　者：梅锴 赵海涛[1] 刘潇然 刘军[1] 熊俊 任保全 魏急波[1] MEI Kai;ZHAO Haitao;LIU Xiaoran;LIU Jun;XIONG Jun;REN Baoquan;WEI Jibo(College of Electronic Science and Technology,National University of Defense Technology,Changsha 410073,China;Institute of Systems Engineering,Military Academy of Sciences,Beijing 100076,China)

机构地区：[1]国防科技大学电子科学学院,湖南长沙410073 [2]军事科学院系统工程研究院,北京100076

出　　处：《通信学报》2022年第1期59-70,共12页Journal on Communications

基　　金：国家自然科学基金资助项目(No.61931020,No.62101569,No.U19B2024,No.62171449,No.62001483);湖南省科技创新计划基金资助项目(No.2020RC2045)。

摘　　要：针对正交频分复用(OFDM)系统,提出一种新型的数据与模型联合驱动下的信道估计算法。该算法结合一种可在线训练的低复杂度学习型估计方法与线性最小均方误差(LMMSE)估计,既赋予信道估计器通过在线训练提升了估计性能的能力,又借助模型解决了在线生成训练数据会造成额外导频开销的问题,提升了系统效率。仿真结果表明,所提算法在低信噪比下的性能和对实际非理想因素的适应性等方面优于传统信道估计算法。For orthogonal frequency division multiplexing(OFDM)systems,a hybrid model and data driven channel estimation algorithm was proposed.Combined with two existing channel estimation methods,including a low complex learning-based channel estimation method and the linear minimum mean square error(LMMSE)channel estimation,the estimator with the ability was facilitated to employ online training to improve estimation performance.Meanwhile,the pilot overhead consumed by generating online training data was saved due to the use of the model-based method in the proposed algorithm,which improved the spectrum efficiency.The simulation results demonstrate that the proposed algorithm has better performance under low signal-to-noise ratio(SNR)and better adaptation to practical imperfections compared with conventional channel estimation methods.

关 键 词：机器学习 数据与模型联合驱动 正交频分复用系统 信道估计 

分 类 号：TN929.5[电子电信—通信与信息系统]