基于子空间投影的并行交替采样系统误差估计

Mismatch estimation for parallel time-interleaved sampling system based on subspace projection

并行交替采样技术可以有效解决单片模数转换器(analog-to-digital converter,ADC)转换速率与量化精度间的矛盾,但是多通道间的失配误差将严重降低系统性能。提出了一种基于子空间投影技术的系统误差估计方法,利用多通道时延对应的频域线性相位矢量与噪声子空间的正交特性实现增益误差以及时基误差的精确估计。该方法迭代次数少,估计精度较高,对噪声以及偏置误差稳健,并且可以同时完成信号重构。仿真数据的处理结果验证了该方法的有效性。

The parallel time-interleaved sampling system can avoid the conflict between sampling rate and quantization precision of a single-chip analog-to-digital converter （ADC）. However, the multi-channel mismatch may degrade the performance of the system greatly. A novel multi-channel sampling system mismatch estima- tion method based on subspace projection technique is presented. The method, which carries out timing and gain mismatch estimation simultaneously by exploiting the properties of the phase differences caused by the time in- terleaved array in frequency domain, is robust to residual offset error and noise. The effectiveness of the pro- posed approach is verified by simulated data.