Frequency-Interleaved ADCs With Adaptive Blind Cyclic Calibration
Frequency-Interleaved ADCs With Adaptive Blind Cyclic Calibration
复制标题
DOI：
10.1109/tim.2020.3003381
复制
发表时间：
2020-12
影响因子：
5.6
通讯作者：
Jinpeng Song;Yu-Hen Hu;Jianping An
中科院分区：
文献类型：
--
作者：
Jinpeng Song;Yu-Hen Hu;Jianping An
提出了一种新的适用于任意通道数频率交错模数转换器（FI-ADC）的数字自适应盲校准技术。它在背景中交替和递归地执行多个固有电路缺陷的估计和校正，包括频谱泄漏、谐波折叠、抖动、I/Q镜像和混叠图像，并且足以通过故意选择的初始猜测来提供投影校准能力。我们认为，这是第一个研究，以解决联合通道识别和误差补偿问题中遇到的FI-ADC设计。为了有效的实现，一个分析的时间交织的等效模型，制定，使得未知的失配参数的数量显着减少，并致力于循环估计的信号被最大限度地抽取。这个衍生模型不仅提出了一个更好的角度理解和解释的FI系统机制，但也使足够的可扩展性和灵活性的开发校准框架。大量的仿真结果进一步表明，这种新的架构实现了计算复杂度和收敛性之间的良好平衡，并提供了一个特殊的和相对平坦的有效位数（ENOB）的性能在很宽的带宽。
A novel digital adaptive blind calibration technique for frequency-interleaved analog-to-digital converters (FI-ADCs) with arbitrary numbers of channels is proposed. It performs the estimation and correction of multiple inherent circuit deficiencies alternately and recursively in the background, including spectral leakage, harmonic folding, jitter, I/Q mirror and aliasing images, and suffices to deliver projected calibration capabilities with deliberately chosen initial guesses. We are of the opinion that this is the first study to address the joint channel identification and error compensation problem encountered in the FI-ADC design. For the efficient implementation, an analytical time-interleaving-like equivalent model is formulated such that the number of unknown mismatch parameters is significantly reduced, and signals devoted to cyclic estimation are maximally decimated. This derived model not only propounds a better perspective for comprehension and interpretation of the FI system mechanism but also enables sufficient scalability and flexibility of the developed calibration framework. Extensive simulation results further show that this novel architecture achieves a good balance between the computational complexity and convergence, and provides an exceptional and relatively flat effective number of bit (ENOB) performance over a wide bandwidth.