Han Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Han Yang has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Yuqi Wang
Institute of Microelectronics of the Chinese Academy of Sciences, Beijing, China
3 shared publications - Ying Hou
Institute of Microelectronics of the Chinese Academy of Sciences, Beijing, China
2 shared publications - Yu Liu
Institute of Microelectronics of the Chinese Academy of Sciences, Beijing, China
2 shared publications - Hao Yang
Institute of Microelectronics of the Chinese Academy of Sciences, Beijing, China
1 shared publication - Jun Zhang
Institute of Microelectronics of the Chinese Academy of Sciences, Beijing, China
1 shared publication - Kang Yu
Institute of Microelectronics of the Chinese Academy of Sciences, Beijing, China
1 shared publication
Latest Publications
- 13 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 30 (6) , 4721-4733
A Phase-Enhanced Neural Network With Dual-Path Transformer for Single-Channel Chest Sound Separation
Auscultation of the chest is a fundamental diagnostic tool for cardiovascular and pulmonary diseases. However, the two main chest sound parts, heart sound (HS) and lung sound (LS), are often mixed, limiting diagnostic accuracy. This paper presents a novel Phase-Enhanced Neural Network (PENN) for HS and LS separation. To address the under-utilization of phase information, PENN integrates a feedforward connection that feeds the input spectrum into the Restorer, enabling phase recovery based on the local inference feature of phase. A time-frequency Dual-Path Transformer (DPT) is employed to expand the network's receptive field and enhance performance. To interpret the effectiveness of PENN, two new metrics, mSI-SDRi and pSI-SDRi, are proposed to separately evaluate the contributions of magnitude and phase. Experiments show that PENN achieves pSI-SDRi improvements of 1.44 dB for HS and 2.25 dB for LS under a LS cutoff frequency ($f_{c\text{lung}}$) of 60Hz. Extensive experimental results demonstrate the effectiveness and robustness of PENN, offering a promising solution to improve the accuracy of auscultation.
- 13 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 30 (6) , 4721-4733
A Phase-Enhanced Neural Network With Dual-Path Transformer for Single-Channel Chest Sound Separation
Auscultation of the chest is a fundamental diagnostic tool for cardiovascular and pulmonary diseases. However, the two main chest sound parts, heart sound (HS) and lung sound (LS), are often mixed, limiting diagnostic accuracy. This paper presents a novel Phase-Enhanced Neural Network (PENN) for HS and LS separation. To address the under-utilization of phase information, PENN integrates a feedforward connection that feeds the input spectrum into the Restorer, enabling phase recovery based on the local inference feature of phase. A time-frequency Dual-Path Transformer (DPT) is employed to expand the network's receptive field and enhance performance. To interpret the effectiveness of PENN, two new metrics, mSI-SDRi and pSI-SDRi, are proposed to separately evaluate the contributions of magnitude and phase. Experiments show that PENN achieves pSI-SDRi improvements of 1.44 dB for HS and 2.25 dB for LS under a LS cutoff frequency ($f_{c\text{lung}}$) of 60Hz. Extensive experimental results demonstrate the effectiveness and robustness of PENN, offering a promising solution to improve the accuracy of auscultation.
- Open Access
- 2 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 118381-118393
A Dual-Mode Compatible CT ADC With FIR DAC and SB Quantization for DSM and IDSM Operations
This paper presents a continuous-time (CT) ADC capable of operating as either a delta-sigma modulator (DSM) or an incremental DSM (IDSM). To address the high jitter sensitivity inherent to oversampled CT systems, an architecture incorporating FIR DACs and a single-bit (SB) quantizer is adopted. A fully numerical loop parameter extraction method based on the least-mean-square (LMS) algorithm is proposed, eliminating the need for tedious algebraic derivations required in conventional approaches. Additionally, a multirate state-space simulation framework is introduced to address compatibility issues arising from excess loop delay (ELD), which conventional approaches fail to adequately handle. At the circuit level, a compact and energy-efficient compensation FIR DAC is implemented based on a configurable capacitive passive summation network. The main FIR DAC with 8 taps employs a Zapped Virtual-Ground-Switched Dual Return-To-Open (DRTO) topology to suppress ISI and the effects of limited reference impedance. Layout asymmetry in the main FIR DAC induces differential-mode (DM) crosstalk, which is alleviated through a layout-level cancellation technique leveraging even-indexed FIR taps. The prototype, fabricated in a TSMC 65 nm LP CMOS process with a core area of 0.235 mm², achieves SNDRpeak of 84.3 dB and 81.3 dB in DSM and IDSM modes, respectively, over a 100 kHz bandwidth. Powered by a single 1.2 V supply, it consumes 269 μW and 277 μW in DSM and IDSM modes, corresponding to FoMSNDR of 170 dB and 166.9 dB.
- 2 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 118381-118393
Open AccessA Dual-Mode Compatible CT ADC With FIR DAC and SB Quantization for DSM and IDSM Operations
This paper presents a continuous-time (CT) ADC capable of operating as either a delta-sigma modulator (DSM) or an incremental DSM (IDSM). To address the high jitter sensitivity inherent to oversampled CT systems, an architecture incorporating FIR DACs and a single-bit (SB) quantizer is adopted. A fully numerical loop parameter extraction method based on the least-mean-square (LMS) algorithm is proposed, eliminating the need for tedious algebraic derivations required in conventional approaches. Additionally, a multirate state-space simulation framework is introduced to address compatibility issues arising from excess loop delay (ELD), which conventional approaches fail to adequately handle. At the circuit level, a compact and energy-efficient compensation FIR DAC is implemented based on a configurable capacitive passive summation network. The main FIR DAC with 8 taps employs a Zapped Virtual-Ground-Switched Dual Return-To-Open (DRTO) topology to suppress ISI and the effects of limited reference impedance. Layout asymmetry in the main FIR DAC induces differential-mode (DM) crosstalk, which is alleviated through a layout-level cancellation technique leveraging even-indexed FIR taps. The prototype, fabricated in a TSMC 65 nm LP CMOS process with a core area of 0.235 mm², achieves SNDRpeak of 84.3 dB and 81.3 dB in DSM and IDSM modes, respectively, over a 100 kHz bandwidth. Powered by a single 1.2 V supply, it consumes 269 μW and 277 μW in DSM and IDSM modes, corresponding to FoMSNDR of 170 dB and 166.9 dB.
- 16 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 72 (1) , 23-27
A 275 μ W 81.5 dB-SNDR and 200 kS/s CT IDSM With an AC-OSA-Based Integrator Using CNRS and FIR DACs
This brief presents an energy-efficient continuous-time incremental delta-sigma modulator (CT IDSM), where the first integrator employs a fourth-order feedforward-compensated amplifier based on an AC-coupled OTA-stacking amplifier (AC-OSA) and utilizes a charge neutralization reset scheme (CNRS). Extending the AC-OSA to incremental structures significantly improves the energy efficiency of the ADC. The core innovation is the use of CNRS for periodic resets, addressing common-mode (CM) disturbances and incomplete differential-mode (DM) signal resets caused by the conventional shorted integration capacitor scheme. Additionally, to enhance area and power efficiency when using high-tap FIR DACs in CT IDSMs, we propose an excess-loop-delay compensation (ELDC) FIR DAC suitable for single-bit (SB) architectures. A third-order single-loop SB CT IDSM with an eight-tap FIR DAC is fabricated in a 65nm CMOS process and achieves SNDR, SNR, and DR values of 81.5dB, 81.8dB, and 85dB, respectively at a Nyquist conversion rate of 200kS/s. The core area is 0.227mm2, and the ADC consumes 275μW. The Schreier (SNDR) figure-of-merit (FoM) is 167.1dB.
- 16 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 72 (1) , 23-27
A 275 μ W 81.5 dB-SNDR and 200 kS/s CT IDSM With an AC-OSA-Based Integrator Using CNRS and FIR DACs
This brief presents an energy-efficient continuous-time incremental delta-sigma modulator (CT IDSM), where the first integrator employs a fourth-order feedforward-compensated amplifier based on an AC-coupled OTA-stacking amplifier (AC-OSA) and utilizes a charge neutralization reset scheme (CNRS). Extending the AC-OSA to incremental structures significantly improves the energy efficiency of the ADC. The core innovation is the use of CNRS for periodic resets, addressing common-mode (CM) disturbances and incomplete differential-mode (DM) signal resets caused by the conventional shorted integration capacitor scheme. Additionally, to enhance area and power efficiency when using high-tap FIR DACs in CT IDSMs, we propose an excess-loop-delay compensation (ELDC) FIR DAC suitable for single-bit (SB) architectures. A third-order single-loop SB CT IDSM with an eight-tap FIR DAC is fabricated in a 65nm CMOS process and achieves SNDR, SNR, and DR values of 81.5dB, 81.8dB, and 85dB, respectively at a Nyquist conversion rate of 200kS/s. The core area is 0.227mm2, and the ADC consumes 275μW. The Schreier (SNDR) figure-of-merit (FoM) is 167.1dB.