单通道小样本信号频率估计算法

Frequency estimation algorithm of small sample single channel signal

针对噪声条件下单通道小样本信号的频率估计问题,提出基于MUSIC方法估计信号频率的算法,通过分析单通道接收信号,结合阵列信号处理方法,根据离散采样间隔和线性阵列阵元间距的关系,提出新的观测数据矩阵构造方法.利用采样数据构造一个Toeplitz矩阵,然后对该矩阵进行特征值分解得到信号子空间和噪声子空间,并通过MUSIC算法实现在单通道较小采样数据量的条件下,精确地估计信号频率.最后经过计算机仿真并与快速傅里叶变换(FFT)算法相比,验证了本文算法的有效性和优越性.

A frequency estimation algorithm based on MUSIC was prososed for signal-channel small sample signal,which analysed the receptive signal and the relation between the discrete sample period and the array element spacing of uniform linear array（ULA）,and then the new Toeplitz matrix could be presented.The signal subspace and noise subspace could be acquired by performing eigenvalue decomposition of the matrix,and the MUSIC algorithm could be used to estimate the signal frequency accurately on the condition of small sample through single channel.Simulation results show that the proposed method is more effective than the fast Fourier transform（FFT）.