一种并行系统时基误差自适应估计方法

Adaptive evaluation method for time-base mismatch of sampling systems

在并行采集系统中,通道间时基延迟的不一致性严重降低了系统性能。针对时间延迟估计算法多基于时域实现,需进行复杂的插值运算以获取采样间隔非整数倍时基延迟的问题,本文基于时基误差的频域模型,将通道间的误差信号建模为自适应滤波器,提出了一种无需插值的估计算法。计算机仿真及实际应用验证结果表明,该方法能动态跟踪时基延迟变化,有效地估计通道时延,具有迭代次数少、运算量小、实时性高的特点。

In parallel time-interleaved data acquisition systems, time-base mismatch between channels degrades the performance of the whole ADC system. Most existing algorithms for time delay estimation are implemented in time domain. When time delay is not an integral multiple of sampling interval, complex interpolation operation is needed, which makes the computation complex and decreases the real-time performance. In this paper, a time-base mismatch adaptive estimation algorithm is proposed by means of modeling mismatch between channels as an adaptive filter in frequency domain. Computer simulations and application results demonstrate that this algorithm could dynamically track the change of time-base mismatch efficiently and avoid the inconvenience of interpolation with the characteristics of less computation and high real-time.