摘要
信道仿真器中不同通道之间的幅相不一致会使系统性能恶化,通常的解决方法是在接收信道中加入校准均衡器,以校正信号传输路径中的幅度和相位失真。为了降低通道间的幅相误差、提高信道均衡效果,文中设计了一种正则化交替频域最小二乘优化算法(Regularized Alternating Frequency Domain Least Squares Optimization,RA-ALO)。该算法采用正则化来确保矩阵求逆时解的正确性,并使用交替最小二乘法(Alternating Least Squares,ALS)寻找最优解来增强均衡器性能。仿真结果表明,在信道均衡方面,与正则化算法和求伪逆算法相比,该算法具有更低的幅相误差、更优的校准性能。
The inconsistency in magnitude and phase between different channels in a channel simulator can deteriorate system performance.The conventional approach to address this issue involves incorporating a calibration equalizer in the receiving channel to conect the magnitude and phase distortions in the signal transmission path.However,the performance of channel equalization is closely tied to the coefficients and structure of the filter.To reduce inter-channel phase errors and enhance channel equalization effectiveness,this paper presents a Regularized Alternating Frequency Domain Least Squares Optimization(RA-ALO)algorithm.This algorithm employs regularization to ensure the accuracy of matrix inversion during solution and utilizes the Alternating Least Squares(ALS)method to seek the optimal solution,thereby enhancing equalizer performance.Simulation results demonstrate that,in terms of channel equalization,compared with regularization algorithms and pseudo inverse algorithms,this algorithm has lower magnitude and phase errors and better calibration performance.
作者
马军
程方
冯姣
张治中
李鹏
赵康
MA Jun;CHENG Fang;FENG Jiao;ZHANG Zhi-zhong;LI Peng;ZHAO Kang(School of Electronics and Information Engineering,Nanjing University of Information Science and Technology,Nanjing 210044,China)
出处
《信息技术》
2025年第12期14-22,共9页
Information Technology
基金
国家重点研发计划资助(2022YFB2902100)
工信部行业专项(TC210H02P)
国家自然基金(61501244,621-71229)。
关键词
信道均衡
幅相误差
正则化
交替最小二乘法
信道仿真器
channel equalization
magnitude phase errors
regularization
Alternating Least Squares
channel simulator
