OAK

Adaptation Methods for DSP-based Receivers

Metadata Downloads
Abstract
이 연구는 DSP 기반 수신기의 적응 방식에서 최첨단 기법에 기여하는 세 가지 TI- ADC 보정 기술과 CTLE 적응을 소개합니다. 제안된 TI-ADC 보정은 오프셋, 이득, 시간 왜곡, 대역폭 불일치를 대상으로 하며, 어떠한 인터리빙 계수에도 적용할 수 있습 니다. 이 방법에는 비귀환 제로 파일럿 신호와 히스토그램 차별화 접근법을 사용하는 불일치 감지가 포함됩니다. 또한, 매개변수 업데이트는 유전 알고리즘과 탐욕적 탐색 접 근법을 포함합니다. 이러한 방법은 하드웨어 요구 사항을 간소화하거나 입력 제한, 전경 신호 해상도와 같은 이전의 한계를 극복함으로써 최첨단 기술에 기여할 수 있습니다. 시뮬레이션 및 측정 결과는 다양한 시스템과 구성에서 상당한 SNDR 향상을 보여주며, 알고리즘의 성능을 입증합니다. 더불어, 소스 저항과 커패시턴스를 동시에 조정할 수 있 는 순차적 CTLE 매개변수 조정 방법이 도입되었으며, 클럭 및 데이터 복구, 자동 이득 제어와같은시스템제어구성요소를최소한의하드웨어오버헤드로효과적으로관리할 수 있습니다.|This work introduces three TI-ADC calibration techniques along with a CTLE adaptation technique, contributing to the state-of-the-art DSP-based receivers adap- tation methods. The proposed TI-ADC calibrations target offset, gain, time skew, and bandwidth mismatches, adaptable to any interleaving factor. The methods include mis- match detection using a non-return-to-zero pilot signal and a histogram comparison approach. Additionally, the parameter updates incorporate a genetic algorithm and a greedy search technique. These methods contribute to the state-of-the-art by either simplifying hardware requirements or overcoming previous limitations, such as input restrictions or foreground signal resolution. Simulations and measurement results show considerable SNDR improvements across various systems and configurations, confirm- ing the algorithms performance. Furthermore, a sequential method for CTLE parame- ter adjustment is introduced, which can concurrently tune both source resistance and capacitance while effectively managing system control components such as clock and data recovery and automatic gain control, with minimal hardware overhead.
Author(s)
Azevedo Tavares, Yang
Issued Date
2025
Type
Thesis
URI
https://scholar.gist.ac.kr/handle/local/18831
Alternative Author(s)
양아제베도타바레스
Department
대학원 전기전자컴퓨터공학부
Advisor
Lee, Minjae
Table Of Contents
Abstract (English) i
Abstract (Korean) iii
List of Contents iv
List of Tables vii
List of Figures viii
1 Introduction 1
2 A Foreground Calibration for M-Channel Time-Interleaved Analog-
to-Digital Converters Based on Genetic Algorithm 7
2.1 Introduction 7
2.2 System Model 9
2.2.1 Limitations on previous state-of-the-art background calibration
methods 13
2.3 Mismatch Magnitude Extraction 15
2.3.1 Foreground Signal 17
2.3.2 Mismatch Extraction 20
2.4 Mismatch Correction 23
2.5 Calibration Parameters Estimation 27
2.6 Results 32
2.6.1 Simulation Result 33
2.6.2 Measurement Result 37
3 A Greedy Search Approach for Time-Interleaved ADCs Calibration
Based on NRZ Input Patterns 45
3.1 Introduction 45
3.2 Proposed Calibration Algorithm 46
3.2.1 Mismatch Estimation and Correction 46
3.2.2 Proposed Mismatch Calibration Convexity 54
3.2.3 Custom Greedy Algorithm 58
3.3 Results 62
3.3.1 Simulation Result 62
3.3.2 Measurement Result 71
4 A Background M-channel Time-Interleaved ADC Calibration for Mul-
tilevel Modulations Based on the Probability Density Function Dif-
ference and a Greedy Algorithm 78
4.1 Introduction 78
4.2 Multilevel Sampling 80
4.2.1 Probability Density Function Intuition 80
4.2.2 ISI Derivation From the Channel Impulse Response 85
4.2.3 Considerations on the Standard TI-ADC Model 88
4.3 Proposed Calibration Algorithm 90
4.3.1 Mismatch Estimation and Correction 90
4.3.2 PDF Difference Estimation 96
4.3.3 Custom Greedy Algorithm 99
4.4 Results 101
4.4.1 Simulation Result 101
4.4.2 Measurement Result 111
5 Advancing Baud Rate Receivers: On Multiple CTLE Parameters Back-
ground Adaptation 119
5.1 Introduction 119
5.2 System Model 122
5.2.1 CTLE Model 127
5.2.2 Channel Model 130
5.2.3 CTLE-ADC-FFE Noise 131
5.3 Proposed CTLE Adaptation 132
5.3.1 Proposed CTLE Adaptation Intuition 132
5.3.2 Single CTLE Adaptation 134
5.3.3 Multiple CTLE Adaptation 141
5.4 Results 147
6 Conclusion 162
References 164
A Derivation of the Generalized TI-ADC SNDR Formula 181
B Modeling and Programming Discussion 184
Acknowledgements 201
Degree
Doctor
Appears in Collections:
Department of Electrical Engineering and Computer Science > 4. Theses(Ph.D)
공개 및 라이선스
  • 공개 구분공개
파일 목록
  • 관련 파일이 존재하지 않습니다.

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.