摘要
数字预失真(Digital Pre-Distortion,DPD)能够提高功率放大器的线性度,并减少非线性失真对信号传输的影响。针对大容量数据高带宽和低信噪比的环境,在系统呈现一定稀疏性时,传统最小均方(Least Mean Square,LMS)无法快速收敛识别功放逆模型。因此,提出一种改进的二范数约束的最小均方(Two Norm Constraint Least Mean Square,2-LMS)算法,通过改变代价函数表达式和步长表达函数来提高算法收敛速度,并减少收敛过程中的稳态误差。此外,引入相邻时刻误差的自相关矩阵,提高了预失真系统的抗噪能力。仿真实验结果表明,基于2-LMS算法的预失真系统在抗噪性能、收敛速度及带外抑制等方面明显优于传统自适应算法,且预失真的邻信道功率比(Adjacent Channel Power Ratio,ACPR)比原基于LMS算法预失真系统优化了-10.3 dB,矢量误差幅度(Error Vector Magnitude,EVM)指标优化了7.5%。
Digital Pre-Distortion(DPD)can improve the linearity of power amplifier and reduce the influence of nonlinear distortion on signal transmission.The traditional Least Mean Square(LMS)cannot quickly converge to identify the inverse model of power amplifier in the environment of large capacity data with high bandwidth and low SNR,and the system presents certain sparseness.In this paper,an Two Norm Constraint Least Mean Square(2-LMS)algorithm with two norm constraints is proposed.By changing the expression of the cost function and the step size expression function,the convergence speed of the algorithm is improved and the steady-state error in the convergence process is reduced.The autocorrelation matrix of the adjacent time error is introduced to improve the anti-noise ability of the predistortion system.Simulation results show that the pre-distortion system based on 2-LMS algorithm is superior to the traditional adaptive algorithm in anti-noise performance,convergence speed and out-of-band suppression,and the pre-distortion of Adjacent Channel Power Ratio(ACPR)is optimized by-10.3 dB and the Vector Error Margin(EVM)index is optimized by 7.5%compared with the original pre-distortion system based on LMS algorithm.
作者
王鑫
李思敏
叶金才
王国富
WANG Xin;LI Simin;YE Jincai;WANG Guofu(Guangxi University of Science and Technology,Liuzhou 545000,China;Guilin University of Electronic Technology,Guilin 541000,China)
出处
《电声技术》
2020年第9期13-17,23,共6页
Audio Engineering
基金
国家自然科学基金资助项目(No.61761009)。
