ADC Nonlinear Equalization Using Background Reference Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital equalization methods for linear or non-linear systems, such as ADCs, face challenges in adapting to changes in analog impairments without interrupting the processing of the main input signal, and require either a purpose-generated training signal or the input signal itself for equalizer adaptation.
Innovation Solution
A method and system that utilize a reference signal added to the input signal to adapt a non-linear equalizer (NLEQ) in the background without stopping the ADC processing, using a simple hardware-generated reference signal and allowing the observation ADC to be slow and relatively non-linear.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a purpose-generated training signal is used to adapt the equalizer, then the equalization accuracy is improved, but the processing of the main ADC input signal is interrupted
Solution Approach 1:
The patent applies preliminary action by continuously adapting the equalizer coefficients in the background using a reference signal before actual equalization is needed. This allows the equalizer to track analog impairments continuously without interrupting main signal processing, as the adaptation准备工作 is done ahead of time in parallel
Solution Approach 2:
The patent segments the signal processing into two independent paths: one for main ADC input signal processing and another for equalizer adaptation using a reference signal. This segmentation allows both operations to proceed simultaneously without interference, resolving the contradiction between accuracy and continuity
2Adaptability or versatility
If the equalizer is updated continuously to track changes in analog impairments, then the adaptability is improved, but the power consumption increases
Solution Approach 1:
The patent introduces a reference signal as an intermediary that carries information about analog impairments. This intermediary allows the equalizer to track impairments indirectly through the reference signal path without requiring continuous high-power processing of the main signal path, thus improving adaptability while controlling power consumption
Solution Approach 2:
The patent uses periodic updating of equalizer coefficients based on reference signal processing rather than continuous updates. This periodic action maintains adaptability by regularly tracking analog impairments while reducing average power consumption compared to continuous full-power processing
3Measurement precision
If a complex calibration circuitry is used to generate reference signal, then the measurement precision is improved, but the device complexity increases
Solution Approach 1:
The patent applies self-service by using the ADC itself to process the reference signal and generate the necessary information for equalizer adaptation. The ADC's own output is used to characterize its impairments, eliminating the need for external complex calibration circuitry while maintaining measurement precision
Solution Approach 2:
The patent creates a simplified copy of the reference signal path that goes through the ADC, allowing impairment characterization without requiring a full complex calibration system. This copying approach provides sufficient accuracy while dramatically reducing device complexity
Data Source
AI summary
A system and method for equalization of a linear or non-linear system. The system includes an adder configured to add an analog reference signal and an input signal, a processing system configured to process a sum of the analog reference signal and the input signal, a non-linear equalizer (NLEQ) configured to process an output of the processing system to remove a distortion incurred by the processing system, a calibration circuitry configured to generate a reconstructed reference signal in digital domain based on measurement of the analog reference signal, and generate coefficients for the NLEQ based on the reconstructed reference signal and the output of the processing system, and a subtractor configured to subtract the reconstructed reference signal from an output of the NLEQ. The analog reference signal may be a sinusoid including single or multiple tones of sinusoids. The non-linear system may be an analog-to-digital converter (ADC).


