• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.11
• /
• pp.1672-1678
• /
• 2022

In this paper, the relay antenna was designed in consideration of the performance of the 28GHz band 5G mobile communication relay horn antenna, such as radiation pattern and return loss. A horn array for 5G mobile communication repeater was designed by arranging the antenna elements in phase, and the performance was analyzed. Unlike conventional WCDMA (3G) and LTE (4G), in millimeter wave band communication, high path loss occurs between transmission and reception. In the design of a 5G millimeter wave horn antenna, antenna performance such as isolation and gain between antenna elements as well as gain and bandwidth of the antenna must be additionally considered. The antenna gain of the single horn antenna (1×1) and the array horn antenna (2×4) in the 28GHz band is about 10.44d Bi and 19.58dBi, respectively, and the return loss is designed to be less than -18dB. It has proven its validity and has been shown to be suitable for application to 5G mobile communication relay system.

• The Bulletin of The Korean Astronomical Society
• /
• v.45 no.1
• /
• pp.36.1-36.1
• /
• 2020

The progenitor of Type Ia supernovae (SNe Ia) is mainly believed to be a carbon/oxygen white dwarf (WD) with non-degenerate (single degenerate) or another WD companion (double degenerate). However, there is little observational evidence of their progenitor system. Recent studies suggest that shock-breakout cooling emission after the explosion can constrain the size of the progenitor system. To do so, we obtained a optical/Near-IR light curve of SN 2019ein, a normal but slightly sub-luminous type Ia supernova, from the very early phase using our high-cadence observation of Intensive Monitoring Survey of Nearby Galaxies (IMSNG). Assuming the expanding fireball model, the simple power-law fitting of the early part of the light curve gives power indices of 1.91 (B) and 2.09 (R) implying radioactive decay of 56Ni is the dominant energy source. By comparison with the expected light curve of the cooling emission, the early observation provides us an upper limit of the companion size of R∗≤1R⊙. This result suggests that we can exclude a large companion such as red giants, which is consistent with the previous study.

• Tuberculosis and Respiratory Diseases
• /
• v.86 no.3
• /
• pp.226-233
• /
• 2023

Background: Inactive or old, healed tuberculosis (TB) on chest radiograph (CR) is often found in high TB incidence countries, and to avoid unnecessary evaluation and medication, differentiation from active TB is important. This study develops a deep learning (DL) model to estimate activity in a single chest radiographic analysis. Methods: A total of 3,824 active TB CRs from 511 individuals and 2,277 inactive TB CRs from 558 individuals were retrospectively collected. A pretrained convolutional neural network was fine-tuned to classify active and inactive TB. The model was pretrained with 8,964 pneumonia and 8,525 normal cases from the National Institute of Health (NIH) dataset. During the pretraining phase, the DL model learns the following tasks: pneumonia vs. normal, pneumonia vs. active TB, and active TB vs. normal. The performance of the DL model was validated using three external datasets. Receiver operating characteristic analyses were performed to evaluate the diagnostic performance to determine active TB by DL model and radiologists. Sensitivities and specificities for determining active TB were evaluated for both the DL model and radiologists. Results: The performance of the DL model showed area under the curve (AUC) values of 0.980 in internal validation, and 0.815 and 0.887 in external validation. The AUC values for the DL model, thoracic radiologist, and general radiologist, evaluated using one of the external validation datasets, were 0.815, 0.871, and 0.811, respectively. Conclusion: This DL-based algorithm showed potential as an effective diagnostic tool to identify TB activity, and could be useful for the follow-up of patients with inactive TB in high TB burden countries.

• Progress in Superconductivity and Cryogenics
• /
• v.25 no.2
• /
• pp.5-9
• /
• 2023

The effect of La_0.7Sr_0.3MnO₃ (LSMO) addition on the superconducting property of Bi_1.6Pb_0.4Sr₂Ca₂Cu₃O_10+δ ((Bi, Pb)-2223) polycrystalline samples was studied. LSMO (0.3 wt.% to 2.0 wt.%) added (Bi, Pb)-2223 samples were prepared by using a solid-state reaction method. The XRD analyses show that as the LSMO addition increases, the volume fraction of the Bi-2223 phase is gradually decreased. The critical temperature (T_c) exhibits a gradual decrease with a single transition as the LSMO amount increases up to 1.0 wt.%, but a further addition of LSMO induces an abrupt decrease of T_c with a dual transition. The analyses on the local structure of the CuO₂ plane from the X-ray absorption fine structure (EXAFS) measurements showed that for the samples with low concentration of LSMO up to 1.0 wt.%, the Cu-O bond length and the CuO₂ plane ordering do not degrade from the values of pure (Bi, Pb)-2223, while they get worsen with a further increase of LSMO addition. These results open up the possibility of LSMO as artificial pinning centers of the (Bi, Pb)-2223 system for power application.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.440-440
• /
• 2022

1970년대부터 집중 건설 된 우리나라의 다목적댐, 홍수조절댐, 용수전용댐 등의 대형 국가 수자원시설물들의 '고령화'가 급속히 진행되어 수리구조물에 대한 안정성을 주기적으로 파악할 수 있는 정밀안전모니터링 체계 구축이 시급한 시점이다. 주기적인 정밀안전모니터링 방법들 중에는 위성 등을 활용한 원격관측 기술들이 최근 시도되고 있다. 위성 영상레이더(SAR; Synthetic Aperture Radar)는 마이크로파 대역의 전자기파를 송·수신하는 능동센서로 날씨 및 주·야간에 영향을 받지 않고 지표면 관측이 가능한 장점이 있다. 특히, 고정산란체 영상레이더 간섭(PSInSAR; Permanent Scatterer Interferometry SAR)기법은 영상레이더 영상에서 긴밀도(coherence)가 상대적으로 높은 수자원시설물과 같은 고정산란체의 위상(phase) 정보를 이용하여 mm급의 측정민감도로 시계열 변위 분석이 가능하다. 또한, 여러 장의 InSAR 영상을 생성하였기 때문에 DEM 오차, 위성궤도 오차, 대기 성분에 의한 지연 오차 등을 보다 정밀하게 제거할 수 있는 장점이 있다 본 연구에서는 국내 중대형 수자원시설물의 정밀안전모니터링을 위하여 고정산란체 영상레이더 간섭 기법을 영암금호방조제, 영주댐, 소양강댐 등에 적용하여 시계열 변위 분석을 수행하였다. 2014년 11월부터 2022년 3월(현재)까지 획득된 Sentinel-1 SLC(Single Look Complex) 위성자료의 상승(Ascending) 궤도 126장 및 하강(Descending)궤도 187장을 각각 활용하였다. 두 위성궤도를 모두 활용하여 수직, 수평 변위 등 3차원 분석을 수행하였으며, 특히 소양강댐 GPS 관측 자료와 정확도 검증에서 연평균 2mm의 RMSE를 보였다. 이를 통해 위성 원격탐사 기술로도 댐, 보, 방조제와 같은 수자원시설물에 대한 시계열 변위 분석을 통한 댐 안전관리가 가능함을 보여주고 있다. 2025년 발사될 국내 C-밴드 SAR 탑재 수자원위성 개발을 통해 한반도 재방문주기를 단축시킴으로써, 한반도 전역의 수자원시설물 정밀안전진단체계 구축이 가능할 것으로 기대된다.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2022.10a
• /
• pp.28-28
• /
• 2022

This paper presents the first genome assemblies of Philippine mangoes that provide valuable reference for varietal improvement and genomic studies on mango and related fruit crops. WE sequenced whole genomes of3 species, Mangifera odorata (Huani), Mangifera altissima (Paho), and Mangifera indica 'Carabao' (Sweet Elena). 'Carabao' is the major export variety of the Philippines; Paho is identified as vulnerable by the IUCN Red List of Threatened Species; Huani has fruit sap acrid which is the primary defense mechanism against insects and birds. We used Falcon, a diploid aware -de novo assembler to assemble SMRT generated long-read sequences. Falcon-unzip was employed to phase the output assembly producing larger contig sets (primary contigs) and shorter contigs corresponding to haplotypes (haplotigs). Assembly statistics were generated by comparing the assembly to a reference genome, Tommy Atkins, using Quality Assessment Tool (QUAST). Moreover, the extent of duplication and completeness of gene content was measured using Benchmarking Universal Single-Copy Orthologs (BUSCO). Draft assemblies with high duplications were processed using Purge Haplotigs and Purge Dups to lessen duplications with minimal impact on genome completeness. De novo assemblies of Huani, Paho and 'Carabao' were then generated with primary contig sizes of 463.64 Mb, 508.95 Mb and 401.51 Mb respectively. These draft assemblies of Huani, Paho and 'Carabao' showed 96.90%, 95.17% and 99.07% complete BUSCOs respectively which is comparable to 'Tommy Atkins' genome (98.6%). Using two mango transcriptome data (pooled RNA-seq from different mango varieties and tissues), 91-96% or 24-30 million reads were successfully mapped back for each generated assembly indicating high degree of completeness. The results obtained demonstrated the highly contiguous, phased, and near complete genome assembly of three Philippine mango species for structural and functional annotation of gene units, especially those with economic importance.

• Journal of Life Science
• /
• v.11 no.2
• /
• pp.111-115
• /
• 2001

This study was designed to clone the SNF2/SW12 helicase-related genes from the fission yeast Schizosaccha-romyces pombe and thereafter to elucidate the common functions of the proteins in this family. The $hrp^{2+}$gene was cloned by polymerase chain reaction amplification using degenerative primers from conserved SNF2 motifs within the ERCC6 gene, which encodes a protein involved in DNA excision repair. Like other SNF2/SW12 family proteins, the deduced amino acid sequence of Hrp2 contains DNA-dependent ATPase/7 helicase domains as well as the chromodomain and the DNA binding domain. This configuration is similar to that of mCHD1 (mouse chromo-ATPase/helicase-DNA-dinding protein 1), suggesting that Hrp2 is a S. pombe homolog of mCHD1, which is thought to function in altering the chromatin structure to control the gene expression. To characterize the function of Hrp2, 4 Uracil-Hrp2 fusion protein, it was purified near homogeneity by affinity chromatography on $Ni^{2+}$-NTA agarose, DEAE-Sepharose ion exchange arid Sephacryl S-200 gel filtration chromatographies. The purified fusion protein exhibited DNA-dependent ATPase activity, which was stimulated by both double-stranded and single-stranded DNA. To determine the steady-state level of $hrp^{2+}$ transcripts during growth, cells were cultured in medium and collected at every 2hr to prepare total RNAs. The northern blot analysis showed that the level of $hrp^{2+}$ transcripts reached its maximum before the cells entered the exponential growth phase and then decreased gradually, This result implies that Hrp2 may be required at early stages of cell growth.h.

• Current Optics and Photonics
• /
• v.7 no.6
• /
• pp.738-744
• /
• 2023

Graphene-based saturable absorbers (SAs) are widely used as laser mode-lockers at various laser oscillators. In particular, transmission-type graphene-SAs with ultrabroad spectral coverage are typically manufactured on transparent substrates with low nonlinearity to minimize the effects on the oscillators. Here, we developed two types of transmitting graphene SAs based on CaF₂ and ZnSe. Using the graphene-SA based on CaF₂, a passively mode-locked mid-infrared Cr:ZnS laser delivers relatively long 540 fs pulses with a maximum output power of up to 760 mW. In the negative net cavity dispersion regime, the pulse width was not reduced further by inhomogeneous group delay dispersion (GDD) compensation. In the same laser cavity, we replaced only the graphene-SA based on CaF₂ with the SA based on ZnSe. Due to the additional self-phase modulation effect induced by the ZnSe substrate with high nonlinearity, the stably mode-locked Cr:ZnS laser produced Fourier transform-limited ~130 fs near 2,340 nm. In the stable single-pulse operation regime, average output powers up to 635 mW at 234 MHz repetition rates were achieved. To our knowledge, this is the first attempt to achieve shorter pulse widths from a polycrystalline Cr:ZnS laser by utilizing the graphene deposited on the substrate with high nonlinearity.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.12
• /
• pp.231-238
• /
• 2023

For boost PFC (Power Factor Correction) converters, various control methods are being studied to achieve unity power factor and low THD (Total Harmonic Distortion) of AC input current. Among them, average current mode control, which controls the average value of the inductor current to follow the current reference, is the most widely used. However, nowadays, as advanced digital control becomes possible with the development of digital processors, predictive control of boost PFC converters is receiving attention. Predictive control is classified into predictive current mode control, which generates duty in advance using a predictive algorithm, and model predictive current control, which performs switching operations by selecting a cost function based on a model. Therefore, this paper simply explains the average current mode control, predictive current mode control, and model predictive current control of the boost PFC converter. In addition, current control under entire load and disturbance conditions is compared and analyzed through simulation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.12
• /
• pp.3383-3397
• /
• 2023

Scene graphs serve as semantic abstractions of images and play a crucial role in enhancing visual comprehension and reasoning. However, the performance of Scene Graph Generation is often compromised when working with biased data in real-world situations. While many existing systems focus on a single stage of learning for both feature extraction and classification, some employ Class-Balancing strategies, such as Re-weighting, Data Resampling, and Transfer Learning from head to tail. In this paper, we propose a novel approach that decouples the feature extraction and classification phases of the scene graph generation process. For feature extraction, we leverage a transformer-based architecture and design an adaptive calibration function specifically for predicate classification. This function enables us to dynamically adjust the classification scores for each predicate category. Additionally, we introduce a Distribution Alignment technique that effectively balances the class distribution after the feature extraction phase reaches a stable state, thereby facilitating the retraining of the classification head. Importantly, our Distribution Alignment strategy is model-independent and does not require additional supervision, making it applicable to a wide range of SGG models. Using the scene graph diagnostic toolkit on Visual Genome and several popular models, we achieved significant improvements over the previous state-of-the-art methods with our model. Compared to the TDE model, our model improved mR@100 by 70.5% for PredCls, by 84.0% for SGCls, and by 97.6% for SGDet tasks.