Samer Medawar

Postcorrection of Pipelined Analog–Digital Converters Based on Input-Dependent Integral Nonlinearity Modeling

The integral nonlinearity (INL) is used for the postcorrection of analog-digital converters (ADCs). An input-frequency-dependent INL model is developed for the postcorrection. The model consists of a static term that is dependent on the ADC output code and a dynamic term that has an additional dependence on the input frequency. The concept of ADC digital output postcorrection by INL is first introduced. The INL model is subtracted from the digital output for postcorrection. The static compensation part is made by adjacent sets of gains and offsets, where each set corrects a range of output codes. The dynamic information, i.e., the frequency dependence of the INL dynamic component is used to construct a set of filter blocks that performs ADC compensation in the time domain. The compensation scheme is applied to the measured data of two ADCs of the same type (Analog Devices AD9430). Performance improvements in terms of spurious-free dynamic range, signal-to-noise and distortion ratio, intermodulation distortion, and noise are obtained. The dynamic compensation part, due to its frequency dependence, can be generalized; hence, a postcorrection block model can be used for compensating multiple ADCs of the same type.

Input-Dependent Integral Nonlinearity Modeling for Pipelined Analog–Digital Converters

Integral nonlinearity (INL) for pipelined analog-digital converters (ADCs) operating at RF is measured and characterized. A parametric model for the INL of pipelined ADCs is proposed, and the corresponding least-squares problem is formulated and solved. The INL is modeled both with respect to the converter output code and the frequency stimuli, which is dynamic modeling. The INL model contains a static and a dynamic part. The former comprises two 1-D terms in ADC code that are a sequence of zero-centered linear segments and a polynomial term. The 2-D dynamic part consists of a set of polynomials whose parameters are dependent on the ADC input stimuli. The INL modeling methodology is applied to simulated and experimental data from a 12-bit commercial ADC running at 210 mega samples per second. It is demonstrated that the developed methodology is an efficient way to capture the INL of nowadays ADCs running at RF, and it is believed that the methodology is powerful for INL-based ADC postcorrection in wideband applications.

Approximate Maximum Likelihood Estimation of Rician K-Factor and Investigation of Urban Wireless Measurements

We consider the problem of estimating the K-factor of a Rician fading wireless channel based on observations of the envelope only, i.e., without phase-information. An approximate Rician power density function (pdf) is introduced to overcome the complexity of the exact pdf. A closed-form maximum-likelihood estimator is derived based on this approximate Rician pdf, while a maximum-likelihood estimator based on the exact Rician pdf appears to be infeasible. An improved estimator is also proposed that features less bias and less variance than the first estimator. The performance of the latter estimator is compared to moment-based estimators that were previously proposed in the literature and is found to have superior performance to closed-form moment estimators, especially for low sample numbers and/or large K-values. The estimator is applied to real wireless macro-cell urban-area measurements. The results show generally low K-factors (below 3dB), with occasional higher values in particular circumstances.

Dynamic Calibration of Undersampled Pipelined ADCs by Frequency Domain Filtering

Integral nonlinearity (INL) is used for the postcorrection of pipeline analog-digital converters (ADCs). An input-frequency-dependent INL model is developed for the compensation. The model consists of a static term that is dependent on the ADC output code, and a dynamic term that has an additional dependence on the input signal frequency. The INL model is subtracted from the digital output for postcorrection. The static compensation is implemented with a look-up-table. The dynamic calibration is performed by a bank of frequency domain filters using an overlap-add structure. Two ADCs of the same type (Analog Devices AD9430) are compensated for in the first three Nyquist bands. The performance improvements in terms of spurious-free dynamic range and intermodulation distortion are investigated. Using the proposed method, improvements up to 17 dB are reported in favorable scenarios.

On the Calibration of Wideband Analog–Digital Converters

The characterization of analog-digital converter (ADC) error is investigated in this letter. More specifically, the ADC static and dynamic errors are clearly identified in a wideband characterization. Improvements obtained by static calibration are presented, and the limitations of dynamic calibration are also investigated.

Static Integral Nonlinearity Modeling and Calibration of Measured and Synthetic Pipeline Analog-to-Digital Converters

The integral nonlinearity (INL) modeling of pipeline analog-to-digital converters (ADCs) is investigated in this paper. The INL is divided into two distinct entities: a low code frequency (LCF) component and a high code frequency (HCF) component. Two static models are developed to represent the INL data. In both models, the LCF component is represented by a low-order polynomial. The HCF modeling is performed using two different basis functions: sinc and Gaussian. The structure of both HCF models is motivated by the pipeline architecture of the ADC under investigation. The model coefficients are estimated by applying the least-squares method to the measured INL data from two samples of a commercial pipeline ADC. The estimated HCF models are compared to each other and to previous models presented in the existing literature. In addition, the modeling methods are applied to synthetic HCF data generated by a pipeline ADC simulation model constructed in MATLAB. The INL models are then used to calibrate the synthetic ADCs, and the improvements in spurious free dynamic range are compared to those obtained when the ADCs are compensated by the INL data. Furthermore, the capability of the HCF modeling to calibrate a given ADC is tested by using the HCF model to compensate a synthetically generated ADC output in which only the measured HCF sequence and noise are added to the quantization process. The results show that the developed HCF models can achieve virtually complete calibration of the considered ADC.

Model order determination and segmentation of analog-digital converters integral non linearity

Analog-digital converter (ADC) integral nonlinearity (INL) modeling is investigated. The model is comprised of two entities: a low code frequency (LCF) component modeled by an L-order polynomial, and a static high code frequency component (HCF), modeled by P linear disjoint segments centered around zero. Both model components are functions of the ADC output code k. A methodical way of estimating the LCF polynomial order L and the set of segments (number of and their borders), is suggested. The method computes the polynomial order L and the set of segments (number and borders) that minimizes the root mean square (RMS) distance between the INL data and its model. The method is applied to measured INL sequences of a 12-bit Analog Devices pipeline ADC (AD9430).

Ricean K-factor estimation and investigation of urban wireless measurements

We consider the problem of estimating the K-factor of a Ricean fading wireless channel based on observations of the envelope only. An approximate Ricean power density function (pdf) is introduced to enable the derivation of a closed-form maximum-likelihood K-factor estimator. A further improved estimator is also proposed, featuring less bias and variance than the first one. The latter estimator outperforms closed form estimators found in the literature especially for low number of samples and/or large K-values. The estimator is applied to real wireless macro-cell urban-area measurements.