针对常规的MUSIC、ESPRIT算法不能估计宽带线性调频(LFM)信号、相干LFM信号和平滑技术解相干在低信噪比下精度不高的问题,提出了一种基于分数阶傅里叶变换(Fractional Fourier Transform,FRFT)与虚拟阵列变换法相结合的宽带相干LFM信号的波达方向(DOA)估计方法。该方法首先将天线阵的观测信号变换到分数阶Fourier域(FRF域),把时域时变的方向向量转化为FRF域时不变的方向向量,然后构造FRF域的阵列信号相关矩阵,用虚拟阵列变换法对相干信号进行解相干,再用ESPRIT算法进行波达方向估计。理论分析和仿真结果表明:在较低信噪比和较少采样快拍时,不仅能实现更准确的DOA估计,而且不损失阵列的孔径,使N个阵元中可估计相干信号源数目达N-1个,提高了阵元数的利用率。
Aiming at the problems that the conventional MUSIC,ESPRIT algorithm can not estimate wideband linear frequency modulation(LFM) signals,and coherent LFM signal and decorrelation precision under low SNR of smoothing techniques is low.A method based on fractional Fourier transform(FRFT) and the virtual array transformation was proposed.By this algorithm,the observed signals were firstly transformed to fractional Fourier domain,and the time-varying direction vector in time domain was changed to the time-invariant vector in fractional Fourier domain,and then constructed the fractional Fourier domain correlation signal matrix of the fractional Fourier domain,and the decorrelation was obtained by virtual array transform.Finally,the DOA was estimated by ESPRIT algorithm.Theoretical analysis and simulation results verified that this approach could achieve more accurate DOA estimates and did not decrease effective array aperture and detected the number of(N-1) coherent signals with N sensors under low SNR and small sample snapshots,which increased utilization of array elements.