Haibin Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Haibin Wang has a total of 127 co-authors. Here are the top co-authors by number of shared publications:
Li Chen
University of Saskatchewan
9 shared publicationsGang Guo
China Institute of Atomic Energy
4 shared publicationsL. Chen
Electrical and Computer Engineering Department, University of Saskatchewan, Saskatoon, SK, Canada
4 shared publicationsXixi Dai
Hohai University
4 shared publicationsIssam Nofal
RoC Technologies
3 shared publicationsLi Cai
China Institute of Atomic Energy
3 shared publications
Latest Publications
- Open Access
Research on Oscillation Suppression Control Method for Contact Scanning Motion
Contact scanning sensors play a crucial role in certain detection equipment, enabling the precise identification and detection of structural degradation to effectively mitigate risks. However, these sensors face a challenge: during high-speed scanning, vibrations can occur due to surface unevenness and friction. These vibrations not only compromise the scanning accuracy but also expedite sensor wear and tear. To address this issue, this study introduces a control strategy. By precisely regulating the actuator’s output, we aim to adjust the sensor’s motion, minimizing vibrations and enhancing the detection system’s stability and accuracy. Simulation results demonstrate that this strategy significantly improves scanning quality and reduces vibrations, validating its effectiveness. These findings pave the way for future research and technological applications.
Open AccessResearch on Oscillation Suppression Control Method for Contact Scanning Motion
Contact scanning sensors play a crucial role in certain detection equipment, enabling the precise identification and detection of structural degradation to effectively mitigate risks. However, these sensors face a challenge: during high-speed scanning, vibrations can occur due to surface unevenness and friction. These vibrations not only compromise the scanning accuracy but also expedite sensor wear and tear. To address this issue, this study introduces a control strategy. By precisely regulating the actuator’s output, we aim to adjust the sensor’s motion, minimizing vibrations and enhancing the detection system’s stability and accuracy. Simulation results demonstrate that this strategy significantly improves scanning quality and reduces vibrations, validating its effectiveness. These findings pave the way for future research and technological applications.
- 23 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 16 (2) , 3001-3012
CFDDR: A Centralized Faulty Data Detection and Recovery Approach for WSN With Faults Identification
One of the most challenging problems in wireless sensor networks (WSNs) is detecting and recovering faulty data. Due to resource limitations, WSNs are subject to many failures, including hardware (permanent faults), software, and communication failures. To ensure data reliability, the data collected by WSNs must be clean of faults. This article presents a centralized faulty data detection and recovery approach, which can detect, recover, and recognize different types of faults (offset, gain, stuck-at, out of bound, and random faults). It operates in two phases: Faulty data detection and recovery and fault type identification (FTI). After detecting the fault, the faulty data will be discarded and replaced by an estimated value based on the Kalman filter. FTI provides a unique solution by determining the types of faults and report them to the end-user for an application-specific decision-making process. The effectiveness of the proposed approach is demonstrated through experimental results, with comparison to state-of-the-art techniques of the same application.
- 23 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 16 (2) , 3001-3012
CFDDR: A Centralized Faulty Data Detection and Recovery Approach for WSN With Faults Identification
One of the most challenging problems in wireless sensor networks (WSNs) is detecting and recovering faulty data. Due to resource limitations, WSNs are subject to many failures, including hardware (permanent faults), software, and communication failures. To ensure data reliability, the data collected by WSNs must be clean of faults. This article presents a centralized faulty data detection and recovery approach, which can detect, recover, and recognize different types of faults (offset, gain, stuck-at, out of bound, and random faults). It operates in two phases: Faulty data detection and recovery and fault type identification (FTI). After detecting the fault, the faulty data will be discarded and replaced by an estimated value based on the Kalman filter. FTI provides a unique solution by determining the types of faults and report them to the end-user for an application-specific decision-making process. The effectiveness of the proposed approach is demonstrated through experimental results, with comparison to state-of-the-art techniques of the same application.
- Open Access
- 29 July 2021
- Published by Institution of Engineering and Technology (IET) in IET Communications
- Vol. 15 (18) , 2285-2299
A reliable and energy efficient dual prediction data reduction approach for WSNs based on Kalman filter
Wireless sensor networks (WSNs) are critically resource‐constrained due to wireless sensor nodes' tiny memory, low processing units, power limitations, and narrow communication bandwidth. The data reduction technique is one of the most widely used techniques to reduce transmitted data over the wireless sensor networks and to minimize the sensor nodes' energy consumption, particularly, the entire network in general. This paper proposes a reliable dual prediction data reduction approach for WSNs. This approach performs data reduction through two phases: the data reduction phase (DRP) and data prediction phase (DPP). The DRP is mainly to decrease the number of transmissions between the sensor node and the sink node, thereby minimizing energy consumption. It also detects faulty data and discards them at the sensor node. The discarded faulty data at the sensor nodes are replaced by estimated values at the sink node to maintain data reliability. DPP runs at the sink node or base station, which works in synchronization with the sensor nodes. This phase is responsible for predicting the non‐transmitted data based on the Kalman filter. The simulation results demonstrate that the proposed approach is efficient and effective in data reduction, data reliability, and energy consumption.
- 29 July 2021
- Published by Institution of Engineering and Technology (IET) in IET Communications
- Vol. 15 (18) , 2285-2299
Open AccessA reliable and energy efficient dual prediction data reduction approach for WSNs based on Kalman filter
Wireless sensor networks (WSNs) are critically resource‐constrained due to wireless sensor nodes' tiny memory, low processing units, power limitations, and narrow communication bandwidth. The data reduction technique is one of the most widely used techniques to reduce transmitted data over the wireless sensor networks and to minimize the sensor nodes' energy consumption, particularly, the entire network in general. This paper proposes a reliable dual prediction data reduction approach for WSNs. This approach performs data reduction through two phases: the data reduction phase (DRP) and data prediction phase (DPP). The DRP is mainly to decrease the number of transmissions between the sensor node and the sink node, thereby minimizing energy consumption. It also detects faulty data and discards them at the sensor node. The discarded faulty data at the sensor nodes are replaced by estimated values at the sink node to maintain data reliability. DPP runs at the sink node or base station, which works in synchronization with the sensor nodes. This phase is responsible for predicting the non‐transmitted data based on the Kalman filter. The simulation results demonstrate that the proposed approach is efficient and effective in data reduction, data reliability, and energy consumption.
- 28 May 2021
- Published by Springer Nature in Applied Physics A
- Vol. 127 (6) , 1-12
Bioinspired micro- and nanostructures used for fog harvesting
- 28 May 2021
- Published by Springer Nature in Applied Physics A
- Vol. 127 (6) , 1-12
Bioinspired micro- and nanostructures used for fog harvesting
- 30 March 2021
- Published by Elsevier in Microelectronics Reliability
- Vol. 120, 114113
Impact of TMR design layouts on single event tolerance in SRAM-based FPGAs
- 30 March 2021
- Published by Elsevier in Microelectronics Reliability
- Vol. 120, 114113
Impact of TMR design layouts on single event tolerance in SRAM-based FPGAs
- 26 February 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Nuclear Science
- Vol. 68 (4) , 394-401
Impact of Single-Event Upsets on Convolutional Neural Networks in Xilinx Zynq FPGAs
Convolutional neural networks (CNNs) are quickly becoming an attractive solution for autonomous vehicles, military weapons, and space exploration. Thanks to their reconfiguration ability, design flexibility, and low power consumption, field-programmable gate arrays (FPGAs) have become a promising candidate for CNN accelerators. However, FPGAs have been proven to be susceptible to radiation-induced single-event upsets (SEUs). One goal of this article is to analyze the impact of quantization on the reliability of CNNs in FPGAs. Therefore, we performed quantization on ZynqNet without affecting its classification accuracy. Meanwhile, we implemented the triple modular redundancy (TMR) version of ZynqNet and we also evaluated the effects of SEUs on these CNNs through both fault injections and neutron exposures. Fault injection results show that TMRed ZynqNet reduces the soft error rate (SER) by 33.59% with a circuit area increase of 111.92% when compared with the standard ZynqNet. The experimental results demonstrate that the quantized ZynqNet reduces the SER by 71.36% with a circuit area reduction of 44.76% when compared with the standard ZynqNet. These results confirm that quantization does contribute to SER reduction of the neural networks. In addition, the operating system on the processing system (PS) side was also found to be highly sensitive to SEUs, and, thus, mitigation techniques should be applied.
- 26 February 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Nuclear Science
- Vol. 68 (4) , 394-401
Impact of Single-Event Upsets on Convolutional Neural Networks in Xilinx Zynq FPGAs
Convolutional neural networks (CNNs) are quickly becoming an attractive solution for autonomous vehicles, military weapons, and space exploration. Thanks to their reconfiguration ability, design flexibility, and low power consumption, field-programmable gate arrays (FPGAs) have become a promising candidate for CNN accelerators. However, FPGAs have been proven to be susceptible to radiation-induced single-event upsets (SEUs). One goal of this article is to analyze the impact of quantization on the reliability of CNNs in FPGAs. Therefore, we performed quantization on ZynqNet without affecting its classification accuracy. Meanwhile, we implemented the triple modular redundancy (TMR) version of ZynqNet and we also evaluated the effects of SEUs on these CNNs through both fault injections and neutron exposures. Fault injection results show that TMRed ZynqNet reduces the soft error rate (SER) by 33.59% with a circuit area increase of 111.92% when compared with the standard ZynqNet. The experimental results demonstrate that the quantized ZynqNet reduces the SER by 71.36% with a circuit area reduction of 44.76% when compared with the standard ZynqNet. These results confirm that quantization does contribute to SER reduction of the neural networks. In addition, the operating system on the processing system (PS) side was also found to be highly sensitive to SEUs, and, thus, mitigation techniques should be applied.
- 28 January 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Device and Materials Reliability
- Vol. 21 (1) , 87-95
Reliable and Radiation-Hardened Push-Pull pFlash Cell for Reconfigured FPGAs
The base component of radiation-hardened flash-based field-programmable gate arrays (FPGAs) is a reconfigurable memory cell, which is reliable and has high driving current. A novel push-pull pflash cell was proposed in this paper, which combined the advantages of indirect coupling and radiation-hardening. The proposed device consisted of two 2T-flash transistors and a pMOS transistor. The push-pull pflash cell was successfully fabricated using 90 nm technology. Through the indirect coupling of the floating gate (FG), the proposed push-pull pflash cell overcame the interference found in sense-switch pflash cells. Compared to conventional sense-switch pflash cells, the proposed push-pull pflash cell had superior driving current, reliability, and radiation-hardening. Band-to-band tunneling-induced hot-electron (BBHE) programming was used to realize ON state, while Fowler-Nordheim (FN) erasing was used to realize OFF state in the proposed device. The electrical properties of both states were characterized. The ON-state driving current and OFF-state leakage current were shown to be consistent. The uniformity of driving current under -1.2 V was below 3%. The proposed device was capable of being programmed/erased cyclically over 10,000 times, and it had a 10-year lifespan with ON/OFF-state stress at 25°C. Additionally, the proposed device was tolerant of total ionizing dose (TID) over 150 Krad(Si). The purposed push-pull pflash cell was suitable for applications in flash-based FPGAs.
- 28 January 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Device and Materials Reliability
- Vol. 21 (1) , 87-95
Reliable and Radiation-Hardened Push-Pull pFlash Cell for Reconfigured FPGAs
The base component of radiation-hardened flash-based field-programmable gate arrays (FPGAs) is a reconfigurable memory cell, which is reliable and has high driving current. A novel push-pull pflash cell was proposed in this paper, which combined the advantages of indirect coupling and radiation-hardening. The proposed device consisted of two 2T-flash transistors and a pMOS transistor. The push-pull pflash cell was successfully fabricated using 90 nm technology. Through the indirect coupling of the floating gate (FG), the proposed push-pull pflash cell overcame the interference found in sense-switch pflash cells. Compared to conventional sense-switch pflash cells, the proposed push-pull pflash cell had superior driving current, reliability, and radiation-hardening. Band-to-band tunneling-induced hot-electron (BBHE) programming was used to realize ON state, while Fowler-Nordheim (FN) erasing was used to realize OFF state in the proposed device. The electrical properties of both states were characterized. The ON-state driving current and OFF-state leakage current were shown to be consistent. The uniformity of driving current under -1.2 V was below 3%. The proposed device was capable of being programmed/erased cyclically over 10,000 times, and it had a 10-year lifespan with ON/OFF-state stress at 25°C. Additionally, the proposed device was tolerant of total ionizing dose (TID) over 150 Krad(Si). The purposed push-pull pflash cell was suitable for applications in flash-based FPGAs.
- 3 April 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Nuclear Science
- Vol. 67 (5) , 805-810
A Low-Overhead FFT Design With Higher SEU Resilience Implemented in FPGA
For area-and-power constraint applications in mission critical systems, the traditional Fast Fourier Transform (FFT) design requires huge hardware resources to do data processing. In this paper, we therefore present a low-overhead radix-2 FFT design that reduces hardware resources by simplifying the first two stages of butterfly computations. Both the proposed design and the traditional design have been implemented on the MicroZus-20 (XC7Z020) platform. Their single event performances have been compared through both fault injections and neutron exposure; and experimental results demonstrate that the proposed design reduces the soft error rate by 53.8%. Also, the proposed FFT reduces hardware resource consumption and circuit area by 64.8% when compared with the traditional design.
- 3 April 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Nuclear Science
- Vol. 67 (5) , 805-810
A Low-Overhead FFT Design With Higher SEU Resilience Implemented in FPGA
For area-and-power constraint applications in mission critical systems, the traditional Fast Fourier Transform (FFT) design requires huge hardware resources to do data processing. In this paper, we therefore present a low-overhead radix-2 FFT design that reduces hardware resources by simplifying the first two stages of butterfly computations. Both the proposed design and the traditional design have been implemented on the MicroZus-20 (XC7Z020) platform. Their single event performances have been compared through both fault injections and neutron exposure; and experimental results demonstrate that the proposed design reduces the soft error rate by 53.8%. Also, the proposed FFT reduces hardware resource consumption and circuit area by 64.8% when compared with the traditional design.
- 24 February 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Device and Materials Reliability
- Vol. 20 (2) , 358-365
The TID Characteristics of a Radiation Hardened Sense-Switch pFLASH Cell
In this paper, the total ionizing dose (TID) effects of the Sense-Switch pFLASH cell consisting of sense (T1) and switch (T2) co-floating gate pFLASH transistors fabricated with a 0.13μm eFLASH technology are studied. Firstly, the impact of TID effects on the programming/erasing state’s characteristics of T1 and T2 transistors are mainly analyzed though Current-Voltage (I-V) and Transconductance-Voltage (Gm-V) curves. Secondly, the re-programming/erasing state’s characteristics of the pFLASH device after TID irradiation, and its threshold voltage, sub-threshold swing, maximum transconductance and other parameters are thoroughly investigated. Lastly, it can be concluded that the threshold voltage of T1 and T2 transistors is mainly determined by the changes of floating gate charge (i.e., programming state: electron emission and hole injection; erasing state: electron injection and hole emission), trap charge concentration in oxide layers around floating gate (Not) and trap charge concentration in the interface trap of tunnel oxide (Nit). Meanwhile, the absolute values of the programming and erasing threshold voltage of T1 and T2 transistors decrease linearly with the cumulative dose, and the change rates are about 2.25 mV/rad (Si) and 1.45 mV/rad (Si), respectively. In addition, the charge change of floating gate under the ionizing radiation can be compensated by re-programming and re-erasing, but the charge of Nit and Not will increase with the increase of cumulative dose.
- 24 February 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Device and Materials Reliability
- Vol. 20 (2) , 358-365
The TID Characteristics of a Radiation Hardened Sense-Switch pFLASH Cell
In this paper, the total ionizing dose (TID) effects of the Sense-Switch pFLASH cell consisting of sense (T1) and switch (T2) co-floating gate pFLASH transistors fabricated with a 0.13μm eFLASH technology are studied. Firstly, the impact of TID effects on the programming/erasing state’s characteristics of T1 and T2 transistors are mainly analyzed though Current-Voltage (I-V) and Transconductance-Voltage (Gm-V) curves. Secondly, the re-programming/erasing state’s characteristics of the pFLASH device after TID irradiation, and its threshold voltage, sub-threshold swing, maximum transconductance and other parameters are thoroughly investigated. Lastly, it can be concluded that the threshold voltage of T1 and T2 transistors is mainly determined by the changes of floating gate charge (i.e., programming state: electron emission and hole injection; erasing state: electron injection and hole emission), trap charge concentration in oxide layers around floating gate (Not) and trap charge concentration in the interface trap of tunnel oxide (Nit). Meanwhile, the absolute values of the programming and erasing threshold voltage of T1 and T2 transistors decrease linearly with the cumulative dose, and the change rates are about 2.25 mV/rad (Si) and 1.45 mV/rad (Si), respectively. In addition, the charge change of floating gate under the ionizing radiation can be compensated by re-programming and re-erasing, but the charge of Nit and Not will increase with the increase of cumulative dose.
- Open Access
- 20 January 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 19490-19503
Soft Error Resilience of Deep Residual Networks for Object Recognition
Convolutional Neural Networks (CNNs) have truly gained attention in object recognition and object classification in particular. When being implemented on Graphics Processing Units (GPUs), deeper networks are more accurate than shallow ones. Residual Networks (ResNets) are one of the deepest CNN architectures used in various fields including safety-critical ones. GPUs have proven to be the major accelerator for CNN models. However, modern GPUs are prone to radiation-induced soft errors, which is a serious issue in safety-compliant systems. In this work, we analyze and propose an approach to address the reliability of ResNet on GPUs. We firstly analyze three popular ResNet models, explicitly, ResNet-50, ResNet-101, and ResNet-152 through NVIDIA’s fault injector, SASSIFI. We perform an in-depth analysis of the model from the perspective of layer and kernel vulnerability. Then, we experimentally show the vulnerability of ResNet models and identify the most vulnerable portions. Finally, we validate our solution, which is a selective-hardening technique, through hardening the worth-hardening kernels to avoid unnecessary overheads. Our strategy is demonstrated to mask up to 93.38% of the injected errors with performance overhead less than 5.35%. Furthermore, the percentage of the errors causing misclassifications can be reduced from 4.2% to 0.104%, thereby significantly improving the model’s reliability.
- 20 January 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 19490-19503
Open AccessSoft Error Resilience of Deep Residual Networks for Object Recognition
Convolutional Neural Networks (CNNs) have truly gained attention in object recognition and object classification in particular. When being implemented on Graphics Processing Units (GPUs), deeper networks are more accurate than shallow ones. Residual Networks (ResNets) are one of the deepest CNN architectures used in various fields including safety-critical ones. GPUs have proven to be the major accelerator for CNN models. However, modern GPUs are prone to radiation-induced soft errors, which is a serious issue in safety-compliant systems. In this work, we analyze and propose an approach to address the reliability of ResNet on GPUs. We firstly analyze three popular ResNet models, explicitly, ResNet-50, ResNet-101, and ResNet-152 through NVIDIA’s fault injector, SASSIFI. We perform an in-depth analysis of the model from the perspective of layer and kernel vulnerability. Then, we experimentally show the vulnerability of ResNet models and identify the most vulnerable portions. Finally, we validate our solution, which is a selective-hardening technique, through hardening the worth-hardening kernels to avoid unnecessary overheads. Our strategy is demonstrated to mask up to 93.38% of the injected errors with performance overhead less than 5.35%. Furthermore, the percentage of the errors causing misclassifications can be reduced from 4.2% to 0.104%, thereby significantly improving the model’s reliability.