연구성과로 돌아가기
2021 연구성과 (141 / 281)
※ 컨트롤 + 클릭으로 열별 다중 정렬 가능합니다.
Excel 다운로드
| WoS | SCOPUS | Document Type | Document Title | Abstract | Authors | Affiliation | ResearcherID (WoS) | AuthorsID (SCOPUS) | Author Email(s) | Journal Name | JCR Abbreviation | ISSN | eISSN | Volume | Issue | WoS Edition | WoS Category | JCR Year | IF | JCR (%) | FWCI | FWCI Update Date | WoS Citation | SCOPUS Citation | Keywords (WoS) | KeywordsPlus (WoS) | Keywords (SCOPUS) | KeywordsPlus (SCOPUS) | Language | Publication Stage | Publication Year | Publication Date | DOI | JCR Link | DOI Link | WOS Link | SCOPUS Link |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ○ | ○ | Article | A Simulation Study on the Effects of Interface Charges and Geometry on Vertical GAA GaN Nanowire MOSFET for Low-Power Application | The effects of interface charges on the performances of gate-all-around (GAA) GaN vertical nanowire MOSFETs with different geometries have been studied. Geometrical effect on the gate current of vertical GAA GaN nanowire MOSFET has also been analysed for the first time. In the ideal condition, the circular geometry nanowire (CGN) MOSFET exhibits the best performance with subthreshold swing (SS) of 62 mV/dec, drain-induced barrier lowering (DIBL) of 14 mV/V, and ON/OFF current ratio (I-ON/I-OFF) of similar to 10(8). The triangular or hexagonal geometry nanowire (TGN or HGN) MOSFET suffer from large gate leakage current due to the field enhancement at sidewall corners. It is also known that interface traps at the sidewall surface of vertical nanowires deteriorate the overall device performance. The HGN MOSFET with m-plane sidewall demonstrates the best performance with SS of 69 mV/dec and DIBL of 13 mV/V, while the TGN MOSFET with a-plane sidewall exhibits the worst performance with SS of 112 mV/dec and DIBL of 101 mV/V. | Thingujam, Terirama; Dai, Quan; Kim, Eunjin; Lee, Jung-Hee | Kyungpook Natl Univ, Sch Elect & Elect Engn, Daegu 41566, South Korea | 57194828103; 57200146404; 57222322113; 57196140713 | jlee@ee.knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0.68 | 2025-07-30 | 6 | 9 | MOSFET; Geometry; Logic gates; Gallium nitride; Gallium arsenide; Performance evaluation; Semiconductor process modeling; Field enhancement; GAA; Gallium Nitride; geometry; interface trap; vertical nanowire | PERFORMANCE; VOLTAGE; TRANSISTORS; FINFET; FET | Field enhancement; GAA; Gallium Nitride; geometry; interface trap; vertical nanowire | Drain current; Gallium nitride; Geometry; III-V semiconductors; Leakage currents; Power MOSFET; Device performance; Different geometry; Drain-induced barrier lowering; Gate-leakage current; Low power application; ON/OFF current ratio; Sub-threshold swing(ss); Vertical nanowires; Nanowires | English | 2021 | 2021 | 10.1109/access.2021.3097367 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | ||||
| ○ | ○ | Article | A Study on Pulse Train Waveforms for High Duty Cycle Sonar Systems: Optimization Scheme and Relationship Between Orthogonality and Bandwidth | Active sonar systems are used in the detection and localization of underwater targets. While traditional approaches use pulsed active sonar (PAS) to transmit short bursts, high duty cycle (HDC) sonar systems have been the focus of recent research and can overcome the shortcomings of PAS. Since HDC sonar systems transmit a long pulse train waveform, we must address sub-pulse interference issues, which requires multiple orthogonal sub-pulses and generalized sinusoidal frequency modulated (GSFM) pulses are suitable for this purpose. Unfortunately, conventional GSFM pulse train design methods do not generate an adequate number of orthogonal sub-pulses. Therefore, we propose an improved GSFM pulse train waveform design approach for overcoming the limitations of these conventional methods. The proposed method is organized into two parts, the first focused on the assessment of the auto-correlation, and the second focused on the optimizing the orthogonality between sub-pulses. Cost function is also carefully designed for this purpose. From simulation experiments, we found that the proposed method was able to produce 30 orthogonal sub-pulses and an optimized waveform that demonstrated better detection performance than conventional waveforms under most conditions. We also verified the performance of the proposed method via actual sea experiments, with analysis showing superior performance matching that of simulations. An analysis on the bandwidth parameter K is also conducted and it is found K D 1 :25 to be the best option. | Kim, Geunhwan; Lee, Kyunkyung; Yoon, Kyungsik; Lee, Seokjin | Kyungpook Natl Univ, Coll IT Engn, Sch Elect Engn, Sch Elect & Elect Engn, Daegu 41566, South Korea; Gimcheon Univ, Dept IT Convergence Engn, Gamcheon 39528, South Korea | 57214800540; 7501515784; 57214791009; 36174416200 | sjlee6@knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0.3 | 2025-07-30 | 5 | 5 | Generalized sinusoidal frequency modulated pulse; high duty cycle sonar; optimization; pulse train waveform | CONTINUOUS ACTIVE SONAR | Generalized sinusoidal frequency modulated pulse; high duty cycle sonar; optimization; pulse train waveform | Bandwidth; Cost functions; Underwater acoustic communication; Active sonar systems; Conventional methods; Detection and localization; Detection performance; Performance matching; Pulse train waveforms; Sinusoidal frequency; Traditional approaches; Sonar | English | 2021 | 2021 | 10.1109/access.2021.3107907 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | ||||
| ○ | ○ | Article | A Time-Slotted Data Gathering Medium Access Control Protocol Using Q-Learning for Underwater Acoustic Sensor Networks | Contention-basedmedium access control (MAC) protocols for underwater acoustic sensor networks are designed to handle packet collisions that are caused by long propagation delays. However, existing protocols are known to suffer from relatively high collisions, which decrease system performance. To enhance system performance, we propose a contention-based MAC protocol that employs a widely-popular machine learning technique, namely, Q-learning. Using Q-learning, the proposed protocol allows the sensor nodes to intelligently select the back-off slots and accordingly schedule the transmission of data packets such that collisions are minimized at the receiver. Unlike in existing protocols, the sensor nodes are not required to exchange scheduling information, which implies that the proposed protocol has low complexity and overhead. Under varying traffic loads and node numbers, the proposed protocol is compared with the state-of-the-art ALOHA-Q for underwater environment (UW-ALOHA-Q), multiple access collision avoidance for underwater (MACA-U) and exponential increase exponential decrease (EIED) protocols. Results demonstrate the effectiveness of the proposed protocol in terms of energy efficiency, channel utilization, and latency. | Ahmed, Faisal; Cho, Ho-Shin | Kyungpook Natl Univ, Sch Elect & Elect Engn, Daegu 41556, South Korea | ; Ahmed, Faisal/MEO-3219-2025 | 57225696280; 35316924900 | hscho@ee.knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 2.12 | 2025-07-30 | 17 | 31 | Media Access Protocol; Protocols; Receivers; Underwater acoustics; Collision avoidance; Schedules; Synchronization; Back-off; collisions; medium access control; machine learning; Q-learning; slot selection; underwater acoustic sensor networks | MAC PROTOCOLS | Back-off; collisions; machine learning; medium access control; Q-learning; slot selection; underwater acoustic sensor networks | Acoustic devices; Energy efficiency; Internet protocols; Learning systems; Reinforcement learning; Sensor networks; Sensor nodes; Underwater acoustic communication; Exponential increase exponential decrease; Machine learning techniques; Medium access control protocols; Multiple access collisions; Scheduling information; Transmission of data; Underwater acoustic sensor networks; Underwater environments; Medium access control | English | 2021 | 2021 | 10.1109/access.2021.3068407 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Article | AEDCN-Net: Accurate and Efficient Deep Convolutional Neural Network Model for Medical Image Segmentation | Image segmentation was significantly enhanced after the emergence of deep learning (DL) methods. In particular, deep convolutional neural networks (DCNNs) have assisted DL-based segmentation models to achieve state-of-the-art performance in fields critical to human beings, such as medicine. However, the existing state-of-the-art methods often use computationally expensive operations to achieve high accuracy and lightweight networks often lack a precise medical image segmentation. Therefore, this study proposes an accurate and efficient DCNN model (AEDCN-Net) based on an elaborate preprocessing step and a resourceful model architecture. The AEDCN-Net exploits bottleneck, atrous, and asymmetric convolution-based residual skip connections in the encoding path that reduce the number of trainable parameters and floating point operations (FLOPs) to learn feature representations with a larger receptive field. The decoding path employs the nearest-neighbor based upsampling method instead of a computationally resourceful transpose convolution operation that requires an extensive number of trainable parameters. The proposed method attains a superior performance in both computational time and accuracy compared to the existing state-of-the-art methods. The results of benchmarking using four real-life medical image datasets specifically illustrate that the AEDCN-Net has a faster convergence compared to the computationally expensive state-of-the-art models while using significantly fewer trainable parameters and FLOPs that result in a considerable speed-up during inference. Moreover, the proposed method obtains a better accuracy in several evaluation metrics compared with the existing lightweight and efficient methods. | Olimov, Bekhzod; Koh, Seok-Joo; Kim, Jeonghong | Kyungpook Natl Univ, Comp Sci & Engn Dept, Daegu 41566, South Korea | ; Olimov, Bekhzod/AAA-9362-2021 | 57220579660; 8958394800; 55138548100 | jhk@knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0.91 | 2025-07-30 | 15 | 15 | Image segmentation; Biomedical imaging; Computational modeling; Convolutional neural networks; Computer architecture; Task analysis; Convolution; Computational efficiency; deep convolutional neural networks; medical image segmentation | U-NET | Computational efficiency; Deep convolutional neural networks; Medical image segmentation | Convolution; Deep neural networks; Digital arithmetic; Image enhancement; Image segmentation; Medical image processing; Floating point operations; Images segmentations; In-field; Learning methods; Learning-based segmentation; Medical image segmentation; Neural network model; Segmentation models; State-of-the-art methods; State-of-the-art performance; Computational efficiency | English | 2021 | 2021 | 10.1109/access.2021.3128607 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Article | Affine Memory Control for Synchronization of Delayed Fuzzy Neural Networks | This paper deals with the synchronization of fuzzy neural networks (FNNs) with time-varying delays. FNNs are more complicated form of neural networks incorporated with fuzzy logics, which provide more powerful performances. Especially, the problem of delayed FNNs's synchronization is of importance in the existence of the network communication. For the synchronization of FNNs with time-varying delays, a novel form of control structure is proposed employing affinely transformed membership functions with memory element. In accordance with affine memory control, appropriate Lyapunov-Krasovskii functional is chosen to design control gain, guaranteeing stability of the systems with delays. Exploiting the more general type of control attributed by affine transformation and memory-type, a novel criterion is derived in forms of linear matrix inequalities (LMIs). As a results, the effectiveness of the proposed control is shown through numerical examples by comparisons with others. | Kwon, Wookyong; Jin, Yongsik; Kang, Dongyeop; Lee, Sangmoon | Elect & Telecommun Res Inst ETRI, Daegu 42995, South Korea; Kyungpook Natl Univ, Sch Elect & Elect Engn, Daegu 41566, South Korea | Lee, Sangmoon/C-4502-2018; Jin, Yongsik/AAH-6959-2021 | 57212541649; 57020309300; 24830558800; 59510733500 | moony@knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0.15 | 2025-07-30 | 1 | 2 | Synchronization; Fuzzy neural networks; Fuzzy control; Delays; Biological neural networks; Linear matrix inequalities; Fuzzy logic; Fuzzy neural networks (FNNs); synchronization; time-varying delay; affine memory control | CHAOS SYNCHRONIZATION; SYSTEMS; STABILIZATION | affine memory control; Fuzzy neural networks (FNNs); synchronization; time-varying delay | Fuzzy inference; Fuzzy logic; Linear matrix inequalities; Linear transformations; Lyapunov functions; Mathematical transformations; Membership functions; Synchronization; System stability; Time delay; Time varying control systems; Affine transformations; Control structure; Fuzzy neural network (FNNs); Lyapunov-Krasovskii functionals; Memory controls; Network communications; Systems with delays; Time varying- delays; Fuzzy neural networks | English | 2021 | 2021 | 10.1109/access.2020.3048170 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Article | An Assembly Quality Inspection System for Bone Conduction Implant Transducers | We designed and implemented a two-axis measurement system to inspect the assembly quality condition of a bone conduction implant (BCI) transducer. The system consists of a laser Doppler vibrometer (LDV), XY manual stage, and two digital scale calipers capable of displaying a position coordinate system. To measure the vibration of the cantilever constituting the transducer vibrational membrane, an XY coordinate system was obtained using FEA software. Based on the derived XY coordinate system, the cantilever vibration displacement of the vibrational membrane was measured for each coordinate using the LDV at 0.5, 0.9, and 2 kHz. To visualize the measured area, we developed a Matlab-based application and then visualized the motion of the cantilever. The alignment and misalignment models of the vibrational membrane and permanent magnet were designed using finite element analysis (FEA) software, and the measured cantilever motions of the vibrational membrane were then compared. Finally, to numerically compare the vibration magnitude of the cantilever, the standard deviation was calculated based on the displacement of each edge of the cantilever. The fabricated BCI transducer had a higher standard deviation (3.5 times at 0.5 kHz, 2.3 times at 0.9 kHz) than the ideally aligned FEA model, but the standard deviation was about eight times lower (at 0.5 and 0.9 kHz) than that of the misaligned case. The results of the numerical comparison indicated that the manufactured BCI transducer was very well assembled. | Shin, Dong Ho; Cho, Hui-Sup | Kyungpook Natl Univ, Inst Biomed Engn Res, Daegu 41944, South Korea; Daegu Gyeongbuk Inst Sci & Technol DGIST, Div Elect & Informat Syst, Daegu 42988, South Korea | 56693502600; 55321104300 | mozart73@dgist.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0 | 2025-07-30 | 0 | 0 | Transducers; Vibrations; Hearing aids; Auditory system; Vibration measurement; Metals; Permanent magnets; Bone conduction implant transducer; vibrational membrane with cantilever structure; assembly quality inspection system; finite element analysis | FLOATING-MASS TRANSDUCER; HISTORY | assembly quality inspection system; Bone conduction implant transducer; finite element analysis; vibrational membrane with cantilever structure | Alignment; Laser Doppler velocimeters; MATLAB; Nanocantilevers; Permanent magnets; Statistics; Transducers; Bone conduction implants; Bone conduction implants (BCI); Cantilever vibrations; Laser Doppler vibrometers; Misalignment models; Numerical comparison; Position coordinates; Vibration magnitude; Vibration analysis | English | 2021 | 2021 | 10.1109/access.2021.3099305 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | ||||
| ○ | ○ | Article | Arithmetic Coding-Based 5-Bit Weight Encoding and Hardware Decoder for CNN Inference in Edge Devices | Convolutional neural networks (CNNs) have gained a huge attention for real-world artificial intelligence (AI) applications such as image classification and object detection. On the other hand, for better accuracy, the size of the CNNs' parameters (weights) has been increasing, which in turn makes it difficult to enable on-device CNN inferences in resource-constrained edge devices. Though weight pruning and 5-bit quantization methods have shown promising results, it is still challenging to deploy large CNN models in edge devices. In this paper, we propose an encoding and hardware-based decoding technique which can be applied to 5-bit quantized weight data for on-device CNN inferences in resource-constrained edge devices. Given 5-bit quantized weight data, we employ arithmetic coding with range scaling for loss-less weight compression, which is performed offline. When executing on-device inferences with underlying CNN accelerators, our hardware decoder enables a fast in-situ weight decompression with small latency overhead. According to our evaluation results with five widely used CNN models, our arithmetic coding-based encoding method applied to 5-bit quantized weights shows a better compression ratio by 9.6x while also reducing the memory data transfer energy consumption by 89.2%, on average, as compared to the case of uncompressed 32-bit floating-point weights. When applying our technique to pruned weights, we obtain better compression ratios by 57.5x-112.2x while reducing energy consumption by 98.3%-99.1% as compared to the case of 32-bit floating-point weights. In addition, by pipelining the weight decoding and transfer with the CNN execution, the latency overhead of our weight decoding with 16 decoding unit (DU) hardware is only 0.16%-5.48% and 0.16%-0.91% for non-pruned and pruned weights, respectively. Moreover, our proposed technique with 4-DU decoder hardware reduces system-level energy consumption by 1.1%-9.3%. | Lee, Jong Hun; Kong, Joonho; Munir, Arslan | Kyungpook Natl Univ, Sch Elect & Elect Engn, Daegu 41566, South Korea; Kyungpook Natl Univ, Sch Elect Engn, Daegu 41566, South Korea; Kansas State Univ, Dept Comp Sci, Manhattan, KS 66506 USA | 57226655958; 25927220400; 24587067400 | joonho.kong@knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0.38 | 2025-07-30 | 1 | 5 | Convolutional neural networks; arithmetic coding; weight compression; edge devices; 5-bit quantization | 5-bit quantization; arithmetic coding; Convolutional neural networks; edge devices; weight compression | Convolution; Data transfer; Decoding; Encoding (symbols); Energy utilization; Image coding; Network coding; Neural networks; Object detection; Quantization (signal); 5-bit quantization; Arithmetic; Arithmetic Coding; Convolutional neural network; Edge device; Encodings; Hardware; Huffman coding; Quantisation; Quantization (signal); Weight compression; Digital arithmetic | English | 2021 | 2021 | 10.1109/access.2021.3136888 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||||
| ○ | ○ | Article | Augmenting Seismic Data Using Generative Adversarial Network for Low-Cost MEMS Sensors | The performance of a deep learning (DL) model depends on sufficient training datasets and its algorithmic structure. Even though seismological research using low-cost micro-electro-mechanical systems (MEMS) sensor received much attention recently, because of the lack of data recorded by such MEMS sensors whose data are usually polluted by different types of noise. Therefore, increasing seismic datasets is required by intelligently generating seismic data through data-augmentation techniques. However, it is difficult to characterize and measure the evolution process of seismic sequences, making the feature extraction and data generation of seismic sequences still a significant challenge. By combining the framework of Generative Adversarial Network (GAN) with long short-term memory (LSTM), attention mechanism and neural network (NN), a novel deep generation model (DGM) named EQGAN is developed to overcome the challenges, which can automatically capture the different time histories and dimension characteristics of seismic sequences, meanwhile stably generating high-quality seismic data. The reality of generated data is qualitatively clarified through the analysis of frequency domain and data autocorrelation distribution. Based on the High-throughput Screening (HTS) Theory, the quantitative evaluation index of statistical metrics is designed, and the generation performance of different machine learning models (standard GAN, LSTM, NN) is compared to prove the stability and effectiveness of EQGAN. The experimental results denote that the EQGAN has excellent stability and performance (up to 81%, much higher than that of other generation models), which provides a suitable data expansion approach for the field of seismological research. | Wu, Aming; Shin, Juyong; Ahn, Jae-Kwang; Kwon, Young-Woo | Kyungpook Natl Univ, Sch Comp Sci & Engn, Daegu 41566, South Korea; Korea Meteorol Adm, Seoul 07062, South Korea | ; Ahn, Jae-Kwang/IQV-7073-2023; Kwon, Young-Woo/HGE-6607-2022 | 58262125900; 57375716300; 57214806947; 57208480210 | ywkwon@knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0.76 | 2025-07-30 | 9 | 11 | Hidden Markov models; Generative adversarial networks; Earthquakes; Data models; Feature extraction; Training; Deep learning; Deep learning; generative adversarial network; data augmentation; Wasserstein distance | KULLBACK-LEIBLER DIVERGENCE; PHASE DETECTION; PICKING | data augmentation; Deep learning; generative adversarial network; Wasserstein distance | Data mining; Earthquakes; Extraction; Frequency domain analysis; Generative adversarial networks; Hidden Markov models; Long short-term memory; MEMS; Seismic response; Seismic waves; Data augmentation; Deep learning; Features extraction; Hidden-Markov models; Low-costs; MEMS (microelectromechanical system); Performance; Seismic sequence; Wasserstein distance; Feature extraction | English | 2021 | 2021 | 10.1109/access.2021.3132901 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Article | Automatic Brittle Fracture Ratio Estimation Using Convolutional Neural Network Regression Based on Classmap Regulation | A convolutional neural network (CNN) based regression is proposed for estimating the brittle fracture ratio (BFR) in a fracture image of a drop weight tear test (DWTT) specimen. Different with the previous complex semantic segmentation-based estimator, the method extracts the feature vector through global average pooling of feature map and calculates the BFR directly through the fully connected layer. By removing decoder network, the number of weights, training time, and required GPU memory dramatically reduced. To train the proposed CNN, a new loss function, which is the sum of L1-norm between class activation map and ground truth inspection image and L1-norm of BFR error, is also designed. To validate the present method, fracture images of 1532, 79, and 158 DWTT specimens obtained from real industrial site were used for training, validation, and test, respectively. The accuracy of the proposed method was evaluated based on the number of test samples with an error of 5% or less divided by the total number of test samples, which is the measure used in real industrial application. Despite having dramatically reduced the number of weights and inference time by 85.8% and 64.8%, respectively, the proposed method has a higher accuracy (96.2%) compared to that of the existing segmentation based BFR estimation method (94.9%). | Jeong, Seung Hyun; Woo, Min Woo; Koo, Gyogwon; Lee, Jong-Hak; Yun, Jong Pil | Korea Inst Ind Technol, Cheonan 31056, South Korea; Kyungpook Natl Univ, Sch Comp Sci & Engn, Daegu 41566, South Korea; Daegu Gyeongbuk Inst Sci & Technol, Daegu 42988, South Korea; Tech Res Labs POSCO, Pohang 37859, South Korea; Univ Sci & Technol, KITECH Sch, Daejeon 34113, South Korea | Lee, Jonghak/G-2764-2012 | 57219224526; 57297761400; 56185419100; 53874024800; 16644164300 | rebirth@kitech.re.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0 | 2025-07-30 | 0 | 0 | Surface cracks; Discrete wavelet transforms; Convolutional neural networks; Feature extraction; Estimation; Training; Inspection; Brittle fracture rate estimator; convolutional neural network regression; drop-weight tear test; heatmap regulation | WEIGHT-TEAR TEST; LOW-TEMPERATURE TOUGHNESS; X70; MICROSTRUCTURE; STEEL | Brittle fracture rate estimator; convolutional neural network regression; drop-weight tear test; heatmap regulation | Convolution; Convolutional neural networks; Drops; Regression analysis; Semantic Segmentation; Semantics; Brittle fracture rate estimator; Convolutional neural network; Convolutional neural network regression; Drop weight tear tests; Fracture rates; Heatmap regulation; Heatmaps; Rate estimator; Ratio estimations; Test specimens; Brittle fracture | English | 2021 | 2021 | 10.1109/access.2021.3117579 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Article | Block-Level Storage Caching for Hypervisor-Based Cloud Nodes | Virtual block devices are heavily used to fulfill the block storage needs of hypervisor-based virtual machine (VM) instances through either local or remote storage spaces. However, a high degree of VM co-location makes it increasingly difficult to physically provision all the necessary block devices using only local storage space. Also, the local storage performance degrades rapidly as workloads interleave. On the other hand, when block devices are acquired through remote storage services, the aggregated network traffic may consume too much cluster-wide network bandwidth in a cloud data center. In order to solve these challenges, we propose a caching scheme for virtual block devices within the hypervisor. The scheme utilizes the physical node's finite local storage space as a block-level cache for the remote storage blocks to reduce the network traffic bound to the storage servers. This allows hypervisor-based compute nodes to serve the hosted VMs' I/O (Input/Output) requests from its local storage as much as possible while enabling VMs to exercise large storage space beyond the capacity of local disks for new virtual disks. Caching virtual disks at block-level in a cloud data center poses several challenges in maintaining high performance while adhering to the virtual disk semantics. We have realized the proposed scheme, called vStore, on Xen hypervisor nodes with factual assessment on its design effectiveness and implementation efficiency. Our comprehensive experimental evaluations show that the proposed scheme substantially reduces the network traffic (49% on average), and incurs less than 12% overheads on the storage I/O performance. | Tak, Byungchul; Tang, Chunqiang; Chang, Rong N.; Seo, Euiseong | Kyungpook Natl Univ, Dept Comp Sci, Daegu 41566, South Korea; Kyungpook Natl Univ, Dept Data Convergence Comp, Daegu 41566, South Korea; IBM Res, Yorktown Hts, NY 10598 USA; IBM Thomas J Watson Res Ctr, Yorktown Hts, NY 10598 USA; Sungkyunkwan Univ, Dept Comp Sci & Engn, Suwon 16419, South Korea | Seo, Euiseong/F-6212-2010 | 6506911621; 8845243400; 7403713275; 14056969000 | euiseong@skku.edu; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0.37 | 2025-07-30 | 2 | 5 | Cloud computing; Servers; Performance evaluation; Virtual machine monitors; Data centers; Bandwidth; Switches; Virtual block device; storage cache; virtualization; network storage | SSD CACHE | network storage; storage cache; Virtual block device; virtualization | Digital storage; Semantics; Virtual machine; Aggregated networks; Cloud data centers; Experimental evaluation; Network bandwidth; Network traffic; Remote storage service; Storage performance; Storage servers; Storage as a service (STaaS) | English | 2021 | 2021 | 10.1109/access.2021.3090308 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Article | Characterizing the Thermal Feasibility of Monolithic 3D Microprocessors | Monolithic 3D (M3D) integration reduces the wire length, which eventually improves energy efficiency and performance compared to 2D integration. However, 3D integration inevitably causes higher on-chip temperature compared to 2D integration due to the increased power density as well as worse heat dissipation. The high on-chip temperature may offset the benefits of the M3D microprocessors due to the following reasons: 1) high on-chip temperature increases leakage power, which degrades energy efficiency. 2) the actual clock frequency is limited at run-time by frequent dynamic thermal management (DTM) invocations. In this paper, for the first time, we explore the thermal feasibility (whether it is possible to achieve high energy efficiency and performance without exceeding threshold temperature) of the M3D microprocessors depending on cooling solutions. For the thermal feasibility study, we construct an integrated framework to investigate the thermal behaviors and thermal feasibility of different types of microprocessors (M3D, 2D, and through-silicon-via based 3D (TSV-3D)) with different cooling solutions. Our thermal-aware evaluation results show that the best configuration of the M3D microprocessors reduces average energy consumption by 27.6% compared to the 2D microprocessor at an iso-frequency (4.0GHz). In addition, at the highest clock frequencies satisfying both design and thermal constraints, the best configuration of the M3D microprocessors improves average system performance by 25.1% and 26.0% compared to the 2D and TSV-3D microprocessors, respectively. | Lee, Ji Heon; Lee, Young Seo; Choi, Jeong Hwan; Amrouch, Hussam; Kong, Joonho; Gong, Young-Ho; Chung, Sung Woo | SK Hynix, DRAM Design Div, Memory Solut Prod Design Grp, Seongnam Si 13558, South Korea; Korea Univ, Dept Comp Sci & Engn, Seoul 02841, South Korea; Univ Stuttgart, Chair Semicond Test & Reliabil, D-70569 Stuttgart, Germany; Kyungpook Natl Univ, Sch Elect Engn, Daegu 41566, South Korea; Kwangwoon Univ, Sch Comp & Informat Engn, Seoul 01897, South Korea | Gong, Young-Ho/AAG-5455-2021; Amrouch, Hussam/AFH-5124-2022; Choi, Jeong/Q-5756-2019 | 57215559728; 57214363991; 57215562016; 37041178500; 25927220400; 55779109700; 7404293097 | yhgong@kw.ac.kr;swchung@korea.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 0.15 | 2025-07-30 | 2 | 2 | Microprocessors; Cooling; Three-dimensional displays; Energy efficiency; System-on-chip; Delays; Clocks; Monolithic 3D integration; on-chip temperature; thermal feasibility; cooling solution | DESIGN | cooling solution; Monolithic 3D integration; on-chip temperature; thermal feasibility | Clocks; Cooling; Electronics packaging; Energy efficiency; Integration; Thermal management (electronics); Three dimensional integrated circuits; 3-D integration; Clock frequency; Cooling solutions; Efficiency and performance; Monolithic 3d integration; Monolithics; On-chip temperature; Silicon via; Thermal; Thermal feasibility; Energy utilization | English | 2021 | 2021 | 10.1109/access.2021.3108628 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Editorial Material | Comments on "ALAM: Anonymous Lightweight Authentication Mechanism for SDN Enabled Smart Homes" | Smart home is intended to be able to enhance home automation systems and achieves goals such as reducing operational costs and increasing comfort while providing security to mobile users. However, an attacker may attempt security attacks in smart home environments because he/she can inject, insert, intercept, delete, and modify transmitted messages over an insecure channel. Secure and lightweight authentication protocols are essential to ensure useful services in smart home environments. In 2020, Iqbal et al. presented an anonymous lightweight authentication protocol for software-defined networking (SDN) enabled smart home, called ALAM. They claimed that ALAM protocol could resist security threats, and also provide secure mutual authentication and user anonymity. This comment demonstrates that ALAM protocol is fragile to various attacks, including session key disclosure, impersonation, and man-in-the-middle attacks, and also their scheme cannot provide user anonymity and mutual authentication. We propose the essential security guidelines to overcome the security flaws of ALAM protocol. | Yu, Sungjin; Das, Ashok Kumar; Park, Youngho | Elect & Telecommun Res Inst, Daejeon 34129, South Korea; Int Inst Informat Technol, Ctr Secur Theory & Algorithm Res, Hyderabad 500032, India; Kyungpook Natl Univ, Sch Elect & Elect Engn, Daegu 41566, South Korea; Kyungpook Natl Univ, Sch Elect Engn, Daegu 41566, South Korea | ; Das, Ashok Kumar/U-2790-2019 | 57203974524; 55450732800; 56962990300 | parkyh@knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 1.36 | 2025-07-30 | 12 | 21 | Cryptanalysis; smart homes; key establishment; authentication; security protocol | USER; SCHEME | Authentication; Cryptanalysis; Key establishment; Security protocol; Smart homes | Ambient intelligence; Automation; Intelligent buildings; Network security; Security systems; Authentication mechanisms; Home automation systems; Lightweight authentication protocols; Man in the middle attacks; Mutual authentication; Security attacks; Security threats; Software defined networking (SDN); Authentication | English | 2021 | 2021 | 10.1109/access.2021.3068723 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Article | Comparison of Deep Learning Techniques for River Streamflow Forecasting | Recently, deep learning (DL) models, especially those based on long short-term memory (LSTM), have demonstrated their superior ability in resolving sequential data problems. This study investigated the performance of six models that belong to the supervised learning category to evaluate the performance of DL models in terms of streamflow forecasting. They include a feed-forward neural network (FFNN), a convolutional neural network (CNN), and four LSTM-based models. Two standard models with just one hidden layer-LSTM and gated recurrent unit (GRU)-are used against two more complex models-the stacked LSTM (StackedLSTM) model and the Bidirectional LSTM (BiLSTM) model. The Red River basin-the largest river basin in the north of Vietnam-was adopted as a case study because of its geographic relevance since Hanoi city-the capital of Vietnam-is located downstream of the Red River. Besides, the input data of these models are the observed data at seven hydrological stations on the three main river branches of the Red River system. This study indicates that the four LSTM-based models exhibited considerably better performance and maintained stability than the FFNN and CNN models. However, the complexity of the StackedLSTM and BiLSTM models is not accompanied by performance improvement because the results of the comparison illustrate that their respective performance is not higher than the two standard models-LSTM and GRU. The findings of this study present that LSTM-based models can reach impressive forecasts even in the presence of upstream dams and reservoirs. For the streamflow-forecasting problem, the LSTM and GRU models with a simple architecture (one hidden layer) are sufficient to produce highly reliable forecasts while minimizing the computation time. | Le, Xuan-Hien; Nguyen, Duc-Hai; Jung, Sungho; Yeon, Minho; Lee, Giha | Kyungpook Natl Univ, Emergency Management Inst, Sangju 37224, South Korea; Thuy Loi Univ, Fac Water Resources Engn, Hanoi 100000, Vietnam; Kyungpook Natl Univ, Dept Adv Sci & Technol Convergence, Sangju 37224, South Korea | ; Nguyen, Hai/AAD-8210-2020; Le, Xuan-Hien/AAZ-9166-2021 | 57209735659; 57215097506; 57209733155; 57223436971; 35069799400 | leegiha@knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 6.51 | 2025-07-30 | 103 | 109 | Rivers; Predictive models; Data models; Floods; Forecasting; Biological system modeling; Computational modeling; Bidirectional LSTM; deep learning; gated recurrent unit; long short-term memory; streamflow forecasting | NEURAL-NETWORK | Bidirectional LSTM; deep learning; gated recurrent unit; long short-term memory; streamflow forecasting | Complex networks; Convolutional neural networks; Deep learning; Feedforward neural networks; Forecasting; Learning systems; Reservoirs (water); Rivers; Stream flow; Watersheds; Complex model; Computation time; Learning techniques; Observed data; Red River basin; Sequential data; Standard model; Streamflow forecasting; Long short-term memory | English | 2021 | 2021 | 10.1109/access.2021.3077703 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Review | Comprehensive Review on Facemask Detection Techniques in the Context of Covid-19 | The outbreak of Coronavirus Disease 2019 (Covid-19) had an enormous impact on humanity. Till May 2021, almost 172 million people have been affected globally due to the contagious spread of Covid-19. Although the distribution of vaccines has been started, the worldwide mass distribution is yet to happen. According to the World Health Organization (WHO), wearing a facemask can reduce the contagious spread of Covid-19 significantly. The governments of different countries have recommended implementing the "no mask, no service" method to impede the spread of Covid-19. However, even the improper wearing of a facemask can obstruct the goal and lead to the spread of the virus. Therefore, to ensure public safety, a system for monitoring facemasks on faces, commonly known as a facemask detection algorithm, is essential for overcoming this crisis. The facemask detection algorithms are part of the object detection algorithms which are used to detect objects in an image. Among the various object detection algorithms, deep learning showed tremendous performance in facemask detection for its excellent feature extraction capability than the traditional machine learning algorithms. However, there remains a lot of scope for future research to build an efficient facemask detection system. Therefore, this study aims to draw attention to the researchers by providing a narrative and meta-analytic review on all the published works related to facemask detection in the context of Covid-19. Because facemask detection algorithms are run mainly by adopting object detection algorithms, this paper also explores the progress of object detection algorithms over the last few decades. A comprehensive analysis of different datasets used in facemask detection techniques by many studies has been explored. The performance comparison among these algorithms is discussed in narrative and meta-analytic approaches. Finally, this study concludes with a discussion of some of the major challenges and future scope in the related field. | Nowrin, Afsana; Afroz, Sharmin; Rahman, Md. Sazzadur; Mahmud, Imtiaz; Cho, You-Ze | Bangladesh Univ Profess, Dept Informat & Commun Technol ICT, Dhaka 1216, Bangladesh; Jahangirnagar Univ, Inst Informat Technol, Dhaka 1342, Bangladesh; Kyungpook Natl Univ, Sch Elect & Elect Engn, Daegu 41566, South Korea | ; Mahmud, Imtiaz/R-1089-2019 | 57226765472; 58256467200; 59860333500; 56203487900; 7404469829 | sazzad@juniv.edu;yzcho@ee.knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 1.3 | 2025-07-30 | 29 | 62 | COVID-19; Feature extraction; Coronaviruses; Detection algorithms; Object detection; Machine learning algorithms; Faces; Covid-19; convolutional neural network; deep neural network; facemask detection; Covid-19 health; object detection; machine learning | OBJECT; CORONAVIRUS | convolutional neural network; Covid-19; Covid-19 health; deep neural network; facemask detection; machine learning; object detection | Deep learning; Feature extraction; Learning algorithms; Object detection; Object recognition; Signal detection; Viruses; Wear of materials; Analytic approach; Comprehensive analysis; Detection algorithm; Extraction capability; Mass distribution; Object detection algorithms; Performance comparison; World Health Organization; Face recognition | English | 2021 | 2021 | 10.1109/access.2021.3100070 | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Article | Deep Learning-Based Automatic Modulation Classification With Blind OFDM Parameter Estimation | Automatic modulation classification (AMC) is an essential factor in dynamic spectrum access to fulfill the spectrum demand of 5G wireless communications for achieving high data rate and low latency. Many deep learning (DL)-based AMC methods have achieved improved accuracy for classifying analog modulation schemes, single-carrier-based modulation schemes, and multi-carrier signals using several DL architectures such as the convolutional neural network (CNN) and long-short term memory (LSTM). However, most conventional DL-based AMC methods have confused the orthogonal frequency multiplexing division (OFDM)-based signals with different OFDM useful symbol lengths. To resolve the issue, we propose a CNN model operating on the fast Fourier transformation window bank (FWB) to extract the useful symbol length in OFDM, which represents the identification of each OFDM-based wireless communication technology. After extracting the OFDM useful symbol length, we propose a DL-based AMC system combined with FWB and in-phase and quadrature-phase signals to classify the OFDM symbol length and single-carrier modulation schemes simultaneously. Furthermore, we explore the constraints of the FWB parameters according to the length and the fast Fourier transformation (FFT) size of the OFDM signal to achieve good classification accuracy through the experiment. We constructed a dataset by generating OFDM signals of different lengths while changing the FFT size in a fixed bandwidth and selecting only quadrature amplitude modulation (QAM) schemes from RadioML2016.10a. Experimental results show the improved classification accuracy by about 30% over conventional classifiers in additive white Gaussian noise, synchronization impairments, and fading environments. | Park, Myung Chul; Han, Dong Seog | Kyungpook Natl Univ, Sch Elect & Elect Engn, Daegu 41566, South Korea | Han, Dong Seog/N-8949-2018 | 56313821300; 7403219442 | dshan@knu.ac.kr; | IEEE ACCESS | IEEE ACCESS | 2169-3536 | 9 | SCIE | COMPUTER SCIENCE, INFORMATION SYSTEMS;ENGINEERING, ELECTRICAL & ELECTRONIC;TELECOMMUNICATIONS | 2021 | 3.476 | 37.9 | 1.59 | 2025-07-30 | 16 | 22 | OFDM; Modulation; Feature extraction; Wireless communication; Synchronization; Quadrature amplitude modulation; Time-domain analysis; Automatic modulation classification; cognitive radio; deep learning; modulation; neural networks; orthogonal frequency-division multiplexing (OFDM) | CARRIER FREQUENCY OFFSET; SIGNAL IDENTIFICATION; SPECTRAL CORRELATION; RECOGNITION; CHANNELS | Automatic modulation classification; cognitive radio; deep learning; modulation; neural networks; orthogonal frequency-division multiplexing (OFDM) | 5G mobile communication systems; Blind equalization; Convolutional neural networks; Deep learning; Fast Fourier transforms; Gaussian noise (electronic); Long short-term memory; Orthogonal frequency division multiplexing; White noise; Additive White Gaussian noise; Automatic modulation classification; Automatic modulation classification (AMC); Fast Fourier transformations; In-phase and quadrature-phase; Quadrature-amplitude modulations (QAM); Single carrier modulation; Wireless communication technology; Modulation | English | 2021 | 2021 | 10.1109/access.2021.3102223 | 바로가기 | 바로가기 | 바로가기 | 바로가기 |
페이지 이동: