• Journal of Communications and Networks
• /
• 제18권3호
• /
• pp.299-310
• /
• 2016

In practical wireless relay communication systems, non-destination nodes are assumed to be idle not receiving signals while the relay sends messages to a particular destination node, which results in reduced bandwidth efficiency. To improve the bandwidth efficiency, we relax the idle assumption of non-destination nodes and assume that non-destination nodes may receive signals from sources. We note that the message relayed to a particular node in such a system gives rise to interference to other nodes. To study such a more general relay system, we consider, in this paper, a relay system in which the relay first listens to the source, then routes the source message to the destination, and finally produces interference to the destination in sending messages for other systems. We obtain capacity upper and lower bounds and study the optimal method to deal with the interference as well as the optimal routing schemes. From analytic results obtained, we find the conditions on which the direct transmission provides higher transmission rate. Next, we find the conditions, by numerical evaluation of the theoretical results, on which it is better for the destination to cancel and decode the interference. Also we find the optimal source power allocation scheme that achieves the lower bound depending on various channel conditions. We believe that the results provided in this paper will provide useful insights to system designers in strategically choosing the optimal routing algorithms depending on the channel conditions.

• 전자공학회논문지
• /
• 제50권8호
• /
• pp.83-93
• /
• 2013

통계적 신호 전송에 관한 위너의 MMSE와 정보이론 관점에서 처음으로 증명한 샤논의 상호 정보량을 비교함으로서 둘 사이의 관련성을 연구하였다. 위너가 신호 전송에서 보려했던 것은 잡음이 존재하는 채널에서 복원하려는 신호의 전송 품질의 근본적 극한(limit)를 계산해내려는 것이라 해석할 수 있다. 반면, 샤논은 엔트로피 개념을 적용하여 상호 정보에 대한 불확실성의 차를 최대화로 하는 전송 정보량의 근본적 상한(upper bound)의 극한(limits)을 계산을 계산하려는 것이라 해석할 수 있다. 본 논문의 관심은 샤논의 점 대 점 채널 용량의 근본적 극한을 계산하는 샤논의 공식을 유도함에 있어 위너의 최소 평균 자승 오차(MMSE) 컴바이너에 의한 최적 위너 필터를 사용했을 때 위너의 MMSE와 샤논의 상호 정보량이 미적분관계로 연결되어 있음을 확인하는 것이며, 위너 필터의 MMSE 와 샤논의 상호 정보량이 만나는 점에서 대역 효율성의 상한과 에너지 효율성의 하한을 계산하였다. mod-${\Lambda}$ AWGN 채널 모델에 의한 적절한 성형 격자 ${\Lambda}$의 선택과 ${\alpha}$의 MMSE 추정에 의해 격자형 부호 방식이 샤논의 원래 채널 용량 극한에 미치게 됨을 확인하였다.

• 건축역사연구
• /
• 제31권6호
• /
• pp.59-68
• /
• 2022

This paper is an architectural historical study on Chimi of Hwangnyongsa Temple. In this research, the shape and cross-section of the chimi are reviewed. The results of the study are as follows. The chimi is a form in which the head part facing the maru is omitted, and the upper and lower body are separated. The upper and lower bonds are assembled into a two types of joint throughout the side of the torso, and then bound with an iron strap. Because of the absence of ridge line in the front, and the narrow curved surface which makes the side plate close to the plane, the entire cross-section is triangular, and the rear plate maintains the shape of the chimi. The naerimmaru connected to the side of the chimi has a slope, so it is clear that the chimi was used on the woojingak-jibung(hipped-roof), and the wing part and back of the chimi are erected on the side roof. The height of the yongmaru and chunyeomaru is about the same and the roofing tiles of those are in contact. The roofing tiles of chunyemaru should be cut to fit the angle of the contacting part. The maru is 30 stories high of roofing tiles as a result of the on-board survey. Based on reference on the shape and timing of the production of chimi, the height of chimi, and the maru is believed to have been built before the Unified Silla Period and used in buildings with at least seven-kan frontage. Buildings corresponding to these construction conditions can be seen as Central hall and East hall in Hwangnyongsa temple.

• 펄프종이기술
• /
• 제39권3호
• /
• pp.70-76
• /
• 2007

The raw oriental lacquer produced in China was used in this study. This lacquer was diluted with ethanol and terpene in different solvent ratios. The diluted lacquer in the solvents had two layers. The upper layer was solvent soluble and the lower contained precipitates. Hanji was treated with the solvent solubles and the mixture containing precipitates, and then properties of the Hanji were evaluated. The average weight, thickness, and density of Hanji treated with oriental lacquer tended to gradually be decreased as the lower concentration of lacquer was used and as the solvent soluble was treated. Tensile strength of the treated Hanji decreased when the diluted lacquer was used and Hanji treated with terpene dilution at the ratio of lacquer to solvent of 1:5 (v/v) showed higher tensile strength than ones treated with ethanol dilution. The folding endurance decreased as the less concentrated lacquer was used and the highest value was obtained when 1:40 (v/v) of lacquer to solvent ratio in both ethanol and terpene mixture. Absorption ratio was higher in the Hanji treated with ethanol mixture, treated with the lacquer in lower concentration, and treated with ethanol mixture. Hanji treated with ethanol dilution at 1:20 (v/v) and the lower ratio of lacquer to solvent showed a hydrophobic property. Amount of the bound dye materials tended to be decreasing as the concentration of lacquer became lower and it was higher in Hanji treated with ethanol mixture than with terpene mixture.

• 정보처리학회논문지B
• /
• 제14B권4호
• /
• pp.287-294
• /
• 2007

실세계의 문제에서 많은 기계학습의 알고리즘들은 데이터의 클래스 불균형 문제에 어려움을 겪는다. 이러한 클래스 불균형 문제를 해결하기 위하여 데이터의 비율을 변경하거나 좀 더 나은 샘플링 전략으로 극복하려는 연구들이 제안되었다. 그러나 데이터의 비율을 변경하는 연구에서는 전체 데이터 분포의 특성을 고려하지 못하고, 샘플링 전략을 제안하는 연구에서는 여러 가지 제한 조건을 고려해야만 한다. 본 논문에서는 위의 두가지 방법의 장점을 모두 포함하는 개선된 집중 샘플링 방법을 제안한다. 제안된 방법에서는 클래스 불균형 문제를 해결하기 위해 학습에 유용한 데이터들을 샘플링하는데 스코어링에 기반한 데이터 분할 방법을 이용한다. 즉, 입력 데이터들에 대해 SOM(Self Organizing Map)의 학습 결과로 얻은 BMU(Best Matching Unit)와의 거리를 계산하고, 이 거리론 스코어라 한다. 측정된 스코어는 오름차순으로 정렬되며, 이 과정에서 입력 데이터의 분포가 재 표현되고, 재 표현된 분포는 전체 데이터의 특성을 대표하게 된다. 그 결과로 얻은 데이터들 중에서 유용하지 못한 데이터들에 대해 제거하는 과정을 수행하여 새로운 학습 데이터 셋을 얻는다. 새로운 학습 데이터 생성 과정에서는 재 표현된 분포의 결과를 두 구간(upper, lower)으로 분할하는데, 두 추간 사이의 데이터들은 유용하지 못한 패턴들로 간주되어 학습에 이용되지 않는다. 본 논문에서 제안한 방법은 클래스 불균형의 비율 감수 훈련 데이터의 크기 감소, 과적합의 방지 등 몇 가지 장점을 보인다. 제안한 방법으로 샘플링된 데이터에 kNN 을 적용하여, 분류 실험한 결과 심한 불균형이 있는 ecoli 데이터의 분류 성능이 최대 2.27배 향상되었다.

• 대한치과보철학회지
• /
• 제37권5호
• /
• pp.698-713
• /
• 1999

Adherence of Candida albicans(C. albicans) to the surface of a denture is believed to be an initial and essential step in the formation of denture-induced stematitis. Previous studies have provided enormous infomation on the relationship between composition of palatine gland/parotid saliva and upper denture stomatitis. Relatively little information is available on the correlation between lower denture stomatitis and sublingual-submandibular ( SLSM ) saliva. The plaque samples were collected from the two sites($100mm^2$) on the inner surface of lower partial denture corresponding to the stematitis and healthy region of the lower partial dentures of 12 denture stomatitis patients and 6 nor-mal persons who wore lower partial dentures. The samples were plated to isolate C. albicans on a selective Saboraud's dextrose agar plate and the isolates were identified by germ tube test and gram staining. The subjects were divided into group I (stomatitis with C. albican), group II (lesion without C. albicans), group III (no lesion but C. albicans), and group IV (normal and healthy denture wearer). Individual SLSM saliva($20{\mu}g$ of protein) was analyzed by SDS-PAGE (SDS -poly-acrylamide gel electrophoresis) with Coomassie brilliant blue and PAS(Periodic Acid Schinff) stain-ing. The salivary proteins separated in the polyacryamide gels were subjected to immunoblot anaysis using anti-lactoferrin, anti-sIgA, and anti-secretory component of sIgA. In this study using custom made acrylic denture resin beads(5mm in diameter) coated with stimulated individual SLSM saliva, the binding ability of individual C. albicans strains to the beads was observed. Levels of C, albicans adhered to the acrylic resin beads were determined by measuring the optical density of the bound C. albicans to the beads at 580nm. The results showed that a higher number of C. albicans was observed in the lesion site than healthy site. The saliva of group I contained more high molecular weight glycoprotein(mucin, MGI) as compared to group II, III and IV. And lactoferrin and sIgA affected to the binding ability of C. albicans to acylic resin beads. Binding ability of individual C. albicans to the acrylic resin coated with respective individual saliva was found to be greater in group I than the other 3 groups. And when bound cells of C. albicans isolated from individual subject #2 to the saliva coated beads were used binding ability of subject #2 saliva coated beads was founed to be greater than the other sutjects. These results suggested that denture induced stomatitis is related to individual patient's salivary protein composition, especially MG-1. Future studies will be directed toward saliva exam-ination of patients who have general disease and analysis of pellicles formed on prosthesis with respect to oral disease.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권7호
• /
• pp.1533-1546
• /
• 2013

Wireless multicast is considered as an effective transmission mode for the future mobile social contact services supported by Long Time Evolution (LTE). Though wireless multicast has an excellent resource efficiency, its performance suffers deterioration from the channel condition and wireless resource availability. Cognitive Radio (CR) and Device to Device (D2D) are two solutions to provide potential resource. However, resource allocation for cognitive wireless multicast based on D2D is still a great challenge for LTE social networks. In this paper, a joint sub-carriers and power allocation model based on D2D for general cognitive radio multicast (CR-D2D-MC) is proposed for Orthogonal Frequency-Division Multiplexing (OFDM) LTE systems. By opportunistically accessing the licensed spectrum, the maximized capacity for multiple cognitive multicast groups is achieved with the condition of the general scenario of imperfect spectrum sensing, the constrains of interference to primary users (PUs) and an upper-bound power of secondary users (SUs) acting as multicast source nodes. Furthermore, the fairness for multicast groups or unicast terminals is guaranteed by setting a lower-bound number of the subcarriers allocated to cognitive multicast groups. Lagrange duality algorithm is adopted to obtain the optimal solution to the proposed CR-D2D-MC model. The simulation results show that the proposed algorithm improves the performance of cognitive multicast groups and achieves a good balance between capacity and fairness.

• Structural Engineering and Mechanics
• /
• 제44권6호
• /
• pp.775-799
• /
• 2012

This work approaches the structural performance of masonry arches that have a small ratio between number of vossoirs and span length. The aim of this research is to compare and validate three different methods of analysis (funicular limit analysis F.L.A., kinematic limit analysis K.L.A. and plane stress Finite Element Analysis F.E.A.) with an experimental campaign. 18 failure tests with arches of different shapes and boundary conditions have been performed. The basic failure mechanism was the formation of enough hinges in the geometry. Nevertheless, in few cases, sliding between vossoirs also played a relevant influence. Moreover, few arches didn't reach the collapse. The FLA and KLA didn't find a solution close to the experimental values for some of the tests. The low number of vossoirs and joints become a drawback for an agreement between kinematic mechanism, equilibrium of forces and geometry constraints. FLA finds a lower bound whereas KLA finds an upper bound of the ultimate load of the arch. FEA is the most reliable and robust method and it can reproduce most of the mechanism and ultimate loads. However, special care is required in the definition of boundary conditions for FEA analysis. Scientific justification of the more suitability of numerical methods in front of classic methods at calculating arches with a few vossoirs is the main original contribution of the paper.

• Earthquakes and Structures
• /
• 제9권3호
• /
• pp.481-501
• /
• 2015

Current design codes and technical recommendations often provide rough indications on how to assess effective stiffness of Reinforced Concrete (R/C) frames subjected to seismic loads, which is a key factor when a linear analysis is performed. The Italian design code (NTC-2008), Eurocode 8 and ACI 318 do not take into account all the structural parameters affecting the effective stiffness and this may not be on the safe side when second-order $P-{\Delta}$ effects may occur. This paper presents a study on the factors influencing the effective stiffness of R/C beams, columns and walls under seismic forces. Five different approaches are adopted and analyzed in order to evaluate the effective stiffness of R/C members, in accordance with the scientific literature and the international design codes. Furthermore, the paper discusses the outcomes of a parametric analysis performed on an actual R/C building and analyses the main variables, namely reinforcement ratio, axial load ratio, concrete compressive strength, and type of shallow beams. The second-order effects are investigated and the resulting displacements related to the Damage Limit State (DLS) under seismic loads are discussed. Although the effective stiffness increases with steel ratio, the analytical results show that the limit of 50% of the initial stiffness turns out to be the upper bound for small values of axial-load ratio, rather than a lower bound as indicated by both Italian NTC-2008 and EC8. As a result, in some cases the current Italian and European provisions tend to underestimate second-order $P-{\Delta}$ effects, when the DLS is investigated under seismic loading.

• 산업경영시스템학회지
• /
• 제46권3호
• /
• pp.186-197
• /
• 2023

In this study, we focus on the improvement of data quality transmitted from a weather buoy that guides a route of ships. The buoy has an Internet-of-Thing (IoT) including sensors to collect meteorological data and the buoy's status, and it also has a wireless communication device to send them to the central database in a ground control center and ships nearby. The time interval of data collected by the sensor is irregular, and fault data is often detected. Therefore, this study provides a framework to improve data quality using machine learning models. The normal data pattern is trained by machine learning models, and the trained models detect the fault data from the collected data set of the sensor and adjust them. For determining fault data, interquartile range (IQR) removes the value outside the outlier, and an NGBoost algorithm removes the data above the upper bound and below the lower bound. The removed data is interpolated using NGBoost or long-short term memory (LSTM) algorithm. The performance of the suggested process is evaluated by actual weather buoy data from Korea to improve the quality of 'AIR_TEMPERATURE' data by using other data from the same buoy. The performance of our proposed framework has been validated through computational experiments based on real-world data, confirming its suitability for practical applications in real-world scenarios.