研究了双基地多载频MIMO雷达中由ADC采样引起的同步误差的3种提取方法。首先建立了同步误差的信号模型,仿真分析了同步误差对测距以及测角的影响,并将其归结为高精度频率估计问题。频谱连续细化法在较小范围内利用连续傅里叶变换来获得高精度的频率估计,既可以降低运算量又能获得较高的频域分辨率;自相关函数法通过对通道分离并补偿阵列延时差后的直达波信号的自相关函数做反正切运算求得同步误差值;MUSIC法首先构造时间同步误差的导向矢量,并利用其向噪声子空间投影,通过搜索空间谱的最大值来得到同步误差值。最后仿真比较了3种方法的性能,结果表明在低信噪比时可采用运算量较大的MUSIC法以取得较好的性能,而在信噪比较高时可采用运算量较小的自相关函数法或者频谱细化法以获得较快的处理速度。
The estimation methods of time synchronization error, which is caused by ADC sampling, in Bistatic multi-cartier-fre- quency MIMO radar are studied. The time synchronization error model is constructed and its influences on range and angle estima- tion are analyzed by simulation. Estimation of synchronization error can be summed up as high-accuracy frequency estimation in a small extent and three estimation methods, which are zoom F~ method, auto-correlation function method and MUSIC method, are studied. The high-accuracy frequency estimation is obtained by continuous fourier transform in a small extent by zoom FFT method which can decrease the computation amounts and obtain high frequency resolution. The synchronization error is calculated by arc tangent operation to the auto-correlation function of direct-wave after the channel separation and time delay difference compensation in auto-correlation function method. In MUSIC method the time synchronization error steering vectors are constructed first and the synchronization error is obtained by searching the maximum value of spatial spectrum which are obtained by projection onto the noise space. Finally their performance are compared and simulated and it can be concluded that the MUSIC method with great com- putational amounts can be used to obtain good performance when the Signal to Noise Ratio(SNR) is low while the zoom FFT meth- od and auto-correlation function method with small computational amounts can be used to obtain rapid p~:ocessing velocity when the SNR is high.