• 한국컴퓨터정보학회논문지
• /
• 제15권6호
• /
• pp.49-56
• /
• 2010

본 논문은 직교주파수분할다중접속방식 기반의 마이크로 기지국을 위한 PHY-MAC 계층에서 최적화된 자원할당 기법에 대해 연구한다. 본 논문에서는 PHY-MAC 계층 간 제어 정보 흐름 및 기능 구현을 포함하여 계층 간 최적화된 2단계 자원할당 기법을 제안한다. 제안하는 2단계 자원할당 기법은 사용자 그룹핑 단계 및 자원할당 단계로 구성된다. 사용자 그룹핑 단계에서는 각 사용자별 PHY-MAC 계층 특성에 따라 사용자들을 매크로 기지국 자원을 할당받을 사용자 그룹과 마이크로 기지국 자원을 할당받을 사용자 그룹으로 분류한다. 자원할당 단계에서는 사용자 그룹 내의 각 사용자별로 실제 할당 받을 자원의 형태 및 양을 결정하는 스케줄러를 정의한다. 제안하는 기법에서는 다이버시티 및 AMC 기법을 스케줄러 정의에 활용한다. 시뮬레이션 결과는 제안하는 방법이 마이크로 기지국을 고려하지 않는 기존 방법에 비해 평균 셀 수율을 40~80% 증가시킬 수 있음을 보인다.

• 방송공학회논문지
• /
• 제2권1호
• /
• pp.8-15
• /
• 1997

본 논문에서는 하나의 TMS320C31 Digital Signal Processor (DSP)를 사용하여 실시간으로 동작하는 Motion Picture Experts Group-1 (MPEG-1) 오디오 인코더 시스템을 구현하였다. 우선 MPEG-1 Audio Layer-2 및 심리음향모델-1 관련 기본 알고리듬을 C-언어로 구현하여 기본 동작을 확인하였다. 그리고 전체실행 시간을 줄이기 위하여, 이를 다시 Texas Instruments (Tl) 어셈블리어로 작성하였다. 마지막으로, MPEG-1 오디오 인코더 시스템을 위한 실제 DSP 하드웨어 회로 보드를 설계, 제작하였다. Analog-to-Digital Converter (ADC) 제어, 입출력 제어, 그리고 DSP 보드에서 PC로의 비트열 전송과 같은 주변 모듈들은 Very High Speed Hardware Description Language (VHDL)을 사용하여 Field Programmable Gate Array (FPGA)로 구현하였다. 제작된 시스템은 48 KHz로 샘플링 되는 스테레오 오디오 신호를 실시간으로 처리하여 192 kbps 비트율로 부호화된 비트열을 출력시킨다. 다양한 형태의 스테레오 오디오 신호를 통해, 제작된 오디오 인코더 시스템의 실시간 동작과 양질의 오디오 신호가 복원됨을 확인하였다.

• 한국환경과학회지
• /
• 제19권2호
• /
• pp.137-148
• /
• 2010

This paper investigates the characteristics of turbulence schemes. Turbulence closures are fundamental for modeling the atmospheric diffusion, transport and dispersion in the boundary layer. In particular, in non-homogeneous conditions, a proper description of turbulent transport in planetary boundary layer is fundamental aspect. This study is based on the Regional Atmospheric Modeling System (RAMS) and combines four different turbulence schemes to assess if the different schemes have a impact on simulation results of vertical profiles. Two of these schemes are Isotropc Deformation scheme (I.Def) and Anisotropic deformation scheme (A.Def) that are simple local scheme based on Smagorinsky scheme. The other two are Mellor-Yamada scheme (MY2.5) and Deardorff TKE scheme (D.TKE) that are more complex non-local schemes that include a prognostic equation for turbulence kinetic energy. The simulated potential temperature, wind speed and mixing ratio are compared against radiosonde observations from the study region. MY2.5 shows consistently reasonable vertical profile and closet to observation. D.TKE shows good results under relatively strong synoptic condition especially, mixing ratio simulation. Validation results show that all schemes consistently underestimated wind speed and mixing ratio but, potential temperature was somewhat overestimated.

• 한국환경과학회지
• /
• 제1권1호
• /
• pp.19-33
• /
• 1992

The high oxidants, which occur the daily maximum concentrations in the afternoon, are transported into the other region via long range transport mechanisms or trapped within the shallow mixing boundary layer and then removed physically (deposition, transport by mountain wind, etc.) and chemically (reaction with local sources). Therefore, modeling formation of photochemical oxidants requires a complex description of both chemical and meteorolog ital processecs . In this study, as a part of air quality studies, we reviewed various aspects of photochemical modeling on the basis of currently available literature. The result of the review shows that the model is based on a set of coupled continuity equations describing advection, diffusion, transport, deposition, chemistry, emission. Also photochemical oxidant models require a large amount of input data concerned with all aspects of the ozone life cycle. First, emission inventories of hydrocarbon and nitrogen oxides, with appropriate spatial and temporal resolution. Second, chemical and photochemical data allowing the quantitative description of the formation of ozone and other photochemically-generated secondary pollutants. Third, dry deposition mechanisms particularly for ozone, PAN and hydrogen peroxide to account for their removal by absorption on the ground, crops, natural vegetation, man-made and water surfaces. Finally, meteorological data describing the transport of primary pollutants away from their sources and of secondary pollutants towards the sensitive receptors where environmental damage may occur. In order to improve our present study, shortcomings and limitation of existing models are pointed out and verification Process through observation is emphasized.

• 정보통신설비학회논문지
• /
• 제2권3호
• /
• pp.36-47
• /
• 2003

Wideband Code Division Multiple Access (WCDMA) is one of the air interface techniques proposed for the third generation (3G) mobile communication system. WCDMA was selected because it fulfills the IMT-2000 requirements for higher data rate trans mission, support of multimedia capabilities and other flexible services due to its variable bit rates and larger bandwidth, improved capacity and coverage, efficient power control and support for advanced and improved detector structures. Performance evaluation of 3G wireless network through simulation plays an important role in the design and implementation of the actual system, aiding the wireless system designer by providing them the necessary performance conformance statistics prior to implementation. In accordance with this goal, a simulator engine was developed entirely on a MATLAB platform to emulate the behaviour of the WCDMA air interface for both the uplink and downlink in a real world fading mobile environment. This paper discuss the development of the simulator along with a brief description of its functionalities and user interface. The WCDMA air interface mode focused in this paper is in accordance to the 3GPPs frequency division duplex (FDD) mode and restricted to the physical layer description. Performance results for the selected cases for the downlink, uplink, varying mobile velocity and sampling rates are also provided.

• Journal of Information Processing Systems
• /
• 제14권1호
• /
• pp.176-190
• /
• 2018

This paper presents a novel convolutional neural network based multi-feature fusion learning method for non-rigid 3D model retrieval, which can investigate the useful discriminative information of the heat kernel signature (HKS) descriptor and the wave kernel signature (WKS) descriptor. At first, we compute the 2D shape distributions of the two kinds of descriptors to represent the 3D model and use them as the input to the networks. Then we construct two convolutional neural networks for the HKS distribution and the WKS distribution separately, and use the multi-feature fusion layer to connect them. The fusion layer not only can exploit more discriminative characteristics of the two descriptors, but also can complement the correlated information between the two kinds of descriptors. Furthermore, to further improve the performance of the description ability, the cross-connected layer is built to combine the low-level features with high-level features. Extensive experiments have validated the effectiveness of the designed multi-feature fusion learning method.

• 한국음향학회지
• /
• 제23권8호
• /
• pp.554-561
• /
• 2004

반도체 기술과 멀티미디어 통신기술이 발달하면서 고품위 영상과 다중 채널의 오디오에 관심을 갖게 되었다. MPEG 오디오 계층 3 디코더는 표준안에 기반을 둔 프로세서로써 기존에 많이 구현되어 있다. MPBG-1오디오 계층3 디코더의 합성필터는 디코더 전체에서 가장 많은 연산을 필요로 하기 때문에 고속 프로세서를 설계하기 위해서는 연산량을 줄일 수 있는 새로운 방식의 합성필터를 필요로 한다. 따라서 본 논문에서는 MPEG-1 오디오 계층 3의 핵심부분인 합성필터 부분을 DALUT (distributed arithmetic look-up table)방식을 이용하여 FPGA (Field Programmable Gate Array)에 구현하였다. 고속 필터를 설계하기 위해서 승산기 대신에 DALUT방식을 사용하였고, 파이프라인 구조를 사용하였으며, 데이터를 코사인 함수와 곱셈한 결과를 테이블로 만듦으로써 곱셈기를 제거하여 30%의 성능향상을 얻었다. 본 논문에서의 하드웨어 설계는 모두 VHDL (VHSIC Hardware Description Language)로 기술하였다. VHDL 시뮬레이션은 ALDEC사의 Active-HDL 6.1과 Model-sim 및 합성은 Synplify Pro 7.2v을 사용하였다. 대상 라이브러리는 XILINX사의 XC4010E, XC4020BX, XC4052 XL, P&R 툴은 XACT Ml.4를 사용하여 구현하였다. 구현된 프로세서는 20MHz∼70MHz사이에서 동작한다.

• 소성∙가공
• /
• 제22권3호
• /
• pp.150-157
• /
• 2013

Fiber metal laminates (FMLs) are layered materials comprised of thin metal sheets and fiber reinforced plastic (FRP). This paper presents the numerical study of the formability enhancement of FMLs composed of an aluminum alloy and self-reinforced polypropylene (SRPP) composite. In this study, a numerical simulation based on finite element (FE) modeling is proposed to evaluate the formability of FMLs using ABAQUS/Explicit. The FE model, which included a single layer of solid and shell elements to model the blank, used discrete layers of the solid element with a contact model and shell elements with a friction based model for the aluminum alloy-composite interface conditions. This method allowed the description of each layer of FMLs and was able to simulate the interaction between the layers. It is noted through this research that the proposed numerical simulation described properly the formability enhancement of the FMLs and the simulation results showed good agreement with experimental results.

• Food Science and Biotechnology
• /
• 제17권2호
• /
• pp.274-278
• /
• 2008

Ethanol extracts from the edible mushroom hinmogi (Tremella fuciformis, TF) were used in the investigation of effects on $\alpha$-glucosidase in vitro and on glucose-uptake in 3T3-L1 mature adipocytes. Addition of the extract significantly inhibited $\alpha$-glucosidase from small intestine of porcine and of rat (about 42 and 35% of control, respectively), and stimulated glucose uptake (about 100% of control), of which activity was higher than that of maitake (Grifola frondosa) X-fraction, a well known anti-diabetic substance. When the ethanol extracts were further partitioned successively by organic solvents and purified by silica gel chromatography, the non-polar layer (F-7) from hexane layer showed highest stimulatory activity of glucose-uptake among layers tested. The major components of the F-7 were 1-monooleoylglycerol and 1-monopalmitoylglycerol. Our report is the first description of TF with stimulatory activity of glucose-uptake. These results suggest that TF extracts may constitute a new source of glucose transport activator and could be employed as a potential anti-diabetic material for treatment and preventing diabetes.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 2006년도 학술대회
• /
• pp.307-312
• /
• 2006

Protection to the multimedia contents is inevitable to ensure that only authorized users be able to access the protected contents for consumption. Since protection mechanisms need to be designed efficiently by exploiting the type of the contents, we propose a protection scheme for the video bitstream encoded by Scalable Video Coding (SVC) technique. Our scheme exploits the property of SVC in which a video is encoded into spatial, temporal, and quality scalability layers. By applying our proposed protection scheme to the appropriate scalability layers we can effectively control the SVC contents completely or partially. Each layer can be flexibly protected with different encryption keys or even with different encryption algorithms. The algorithms that are used to protect each layer are described by the standardized protection description tool, which is the MPEG-21 Intellectual Property Management and Protection (IPMP) Components. In this paper, we present the design of the proposed layered SVC protection scheme, its implementation and experimental results. The experiment result shows that the proposed layered SVC protection scheme is very effective and can easily be applied.