• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.12
• /
• pp.1-12
• /
• 1998

In this paper, a design and implementation of VHDL-to-C mapping in the VHDL compiler back-end is described. The analyzed data in an intermediate format(IF), produced by the compiler front-end, is transformed into a C-code model of VHDL semantics by the VHDL-to-C mapper. The C-code model for VHDL semantics is based on a functional template, including declaration, elaboration, initialization and execution parts. The mapping is carried out by utilizing C mapping templates of 129 types classified by mapping units and functional semantics, and iterative algorithms, which are combined with terminal information, to produce C codes. In order to generate the C program, the C codes are output to the functional template either directly or by combining the higher mapping result with intermediate mapping codes in the data queue. In experiments, it is shown that the VHDL-to-C mapper could completely deal with the VHDL analyzed programs from the compiler front-end, which deal with about 96% of major VHDL syntactic programs in the Validation Suite. As for the performance, it is found that the code size of VHDL-to-C is less than that of interpreter and worse than direct code compiler of which generated code is increased more rapidly with the size of VHDL design, and that the VHDL-to-C timing overhead is needed to be improved by the optimized implementation of mapping mechanism.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.12
• /
• pp.109-120
• /
• 1998

An image segmentation approach based on the fuzzy worm searching and MIN-MAX clustering algorithm is proposed in this paper. This algorithm deals with fuzzy worm value and min-max node at a gross scene level, which investigates the edge information including fuzzy worm action and spatial relationship of the pixels as the parameters of its objective function. But the conventional segmentation methods for edge extraction generally need the mask information for the algebraic model, and take long run times at mask operation, whereas the proposed algorithm has single operation according to active searching of fuzzy worms. In addition, we also propose both genetic fuzzy worm searching and genetic min-max clustering using genetic algorithm to complete clustering and fuzzy searching on grey-histogram of image for the optimum solution, which can automatically determine the size of ranges and has both strong robust and speedy calculation. The simulation results showed that the proposed algorithm adaptively divided the quantized images in histogram region and performed single searching methods, significantly alleviating the increase of the computational load and the memory requirements.

• Proceedings of the Korean Society of Applied Pharmacology
• /
• 1992.05a
• /
• pp.64-64
• /
• 1992

1. Eudragit RS, RL, E, S 및 L을 이용하여 유중건조법으로 제조한 $\beta$-락탐계 항생제 (Amoxicillin, Cephalexin) 마이크로캅셀 모두가 구상 성형성과 정립성이 양호한 결과를 얻었다. 2. 분리 분산제 (aluminium tristearate)의 양을 고정시켰을 때에는 일정한 입자 분포도를 가지는 마이크로캅셀을 얻을 수 있었고, 양을 증가시켰을 때에는 입자 크기는 작아 졌으며, 용출율은 증가하였다. 3. Eudragit RS 및 S 마이크로캅셀제제로부터의 약물용출은 저조하였고, Eudragit RL 및 L 제제로부터의 약물용출은 양호하였다. 따라서 Eudragit RS/RL 마이크로캅셀제제에서는 Eudragit RL의 양을 증가시킬수록 약물의 용출율이 증가하였으며, Eudragit S/L 마이크로캅셀제제에서는 Eudragit L의 양을 증가시킬수록 약물의 용출율이 증가하였다. 약물방출 실험결과, Amoxicillin 함유Eudragit RS/RL (25/75) 마이크로캅셀, cephalxin함유 Eudragit RS/RL (75/25) 마이크로캅셀 및 Eudragit S/L (75/25) 마이크로캅셀제제는 유용한 제제로 펑가되었다. 또한 수용성 고분자인 polyethylene glycol을 혼합하어 제조하는 것에 의해 방출조절성 마이크로캅셀의 제조가 가능하였다. 유중건조법을 이용하여 본 연구방법으로 저조한 마이크로캅셀제제는 투어횟수를 줄일 수 있을 뿐만 아니라 생체에 대해 안전하고 재현성에 확보되는 유용한 제제로 판단된다. 앞으로 연구를 계속 수행하여, 특히 약물의 물리화학적 성질 및 생제내 투어 후의 생물약제학적 펑가를 엄격히 하므로써 안정성이 심히 문제시되는 다용 약물계열에 대한 조절방출성제제의 개발을 기대할 수 있다.은 해리항수의 역수이므로 해리항수가 적을수록 $\beta$ 수용체에 대한 친화력이 큰 약물이다. 시사되었으며, 이 조직에서 또한 5-$HT_2$와 5-$HT_3$ 수용체의 존재를 확인하고 각각의 기능을 분명히 했다.가 수월하게 하였고 메모리를 동적으로 관리할 수 있게 하였다. 또한 기존의 smpl에 디버깅용 함수 및 설비(facility) 제어용 함수를 추가하여 시뮬레이션 프로그램 작성을 용이하게 하였다. 예를 들면 who_server(), who_queue(), pop_Q(), push_Q(), pop_server(), push_server(), we(), wf(), printfct() 같은 함수들이다. 또한 동시에 발생되는 사건들의 순서를 조종하기 위해, 동시에 발생할 수 있는 각각의 사건에 우선순위를 두어 이 우선 순위에 의하여 사건 리스트(event list)에서 자동적으로 사건들의 순서가 결정되도록 확장하였으며, 설비 제어방식에 있어서도 FIFO, LIFO, 우선 순위 방식등을 선택할 수 있도록 확장하였다. SIMPLE는 자료구조 및 프로그램이 공개되어 있으므로 프로그래머가 원하는 기능을 쉽게 추가할 수 있는 장점도 있다. 아울러 SMPLE에서 새로이 추가된 자료구조와 함수 및 설비제어 방식등을 활용하여 실제 중형급 시스템에 대한 시뮬레이션 구현과 시스템 분석의 예를 보인다._3$", chain segment, with the activation energy of carriers from the shallow trap with 0.4[eV], in he amorphous regions.의 증발산율은 우기의 기상자료를 이용하여 구한 결과 0.05 - 0.10 mm/hr 의 범위로서 이로 인한 강우손실량은 큰 의미가 없음

• Journal of the Microelectronics and Packaging Society
• /
• v.10 no.2
• /
• pp.7-12
• /
• 2003

The $Bi_{3.3}La_{0.7}O_{12}$(BLT) capacitors with Metal-Ferroelectric-Insulator-Silicon structure were prepared on $SiO_2/Si$ substrates by using sol-gel method. The BLT thin films annealed at $650^{\circ}C$ and $700^{\circ}C$ showed randomly oriented perovskite crystalline structures. The full with at half maximum (FWHM) of the (117) main peak was decreased from $0.65^{\circ}$ to $0.53^{\circ}$ with increasing the annealing temperature from $650^{\circ}C$ to $700^{\circ}C$, indicating the improvement in the crystalline quality of the film. In addition, the grain size and $R_rms$ , values were increased with increasing the annealing temperatures, showing the rough film surface at higher annealing temperatures. From the capacitance-voltage (C-V) measurements, the memory window voltage of the BLT film annealed at $700^{\circ}C$ was found to be about 0.7 V at an applied voltage of 5 V. The leakage current density of the BLT film annealed at $700^{\circ}C$ was about $3.1{\times}10^{-8}A/cm^2$.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.45 no.2
• /
• pp.118-124
• /
• 2008

For forward error correction, DVB-S2, which is the digital video broadcasting forward error coding and modulation standard for satellite television, uses a system based the concatenation of BCH with LDPC inner coding. In DVB-S2 the LDPC codes are defined for 11 different code rates, which means that a DVB-S2 LDPC decoder should support multiple code rates. Seven of the 11 code rates, 3/5, 2/3, 3/4, 4/5, 5/6, 8/9, and 9/10, are regular and the rest four code rates, 1/4, 1/3, 2/5, and 1/2, are irregular. In this paper we propose a flexible decoder for the regular LDPC codes. We combined the partially parallel decoding architecture that has the advantages in the chip size, the memory efficiency, and the processing rate with Benes network to implement a DVB-S2 LDPC decoder that can support multiple code rates with a block size of 64,800 and can configure the interconnection between the variable nodes and the check nodes according to the parity-check matrix. The proposed decoder runs correctly at the frequency of 200MHz enabling 193.2Mbps decoding throughput. The area of the proposed decoder is $16.261m^2$ and the power dissipation is 198mW at a power supply voltage of 1.5V.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37C no.11
• /
• pp.1170-1181
• /
• 2012

Wireless Sensor Networks(WSNs) are used in diverse applications. In general, sensor nodes that are easily deployed on specific areas have many resource constrains such as battery power, memory sizes, MCUs, RFs and so on. Hence, first of all, the efficient energy consumption is strongly required in WSNs. In terms of event states, event-driven deliverly model (i.e. surveillance and reconnaissance applications) has several characteristics. On the basis of such a model, clustering algorithms can be mostly used to manage sensor nodes' energy efficiently owing to the advantages of data aggregations. Since a specific node collects packets from its child nodes in a network topology and aggregates them into one packet to relay them once, amount of transmitted packets to a sink node can be reduced. However, most clustering algorithms have been designed without considering can be reduced. However, most clustering algorithms have been designed without considering characteristics of event-driven deliverly model, which results in some problems. In this paper, we propose enhanced clustering algorithms regarding with both targets' movement and energy efficiency in order for applications of surveillance and reconnaissance. These algorithms form some clusters to contend locally between nodes, which have already detected certain targets, by using a method which called CHEW (Cluster Head Election Window). Therefore, our proposed algorithms enable to reduce not only the cost of cluster maintenance, but also energy consumption. In conclusion, we analyze traces of the clusters' movements according to targets' locations, evaluate the traces' results and we compare our algorithms with others through simulations. Finally, we verify our algorithms use power energy efficiently.

• Journal of KIISE:Databases
• /
• v.35 no.6
• /
• pp.505-517
• /
• 2008

This paper proposes a rewriting algorithm of OWL-DL ontology query in SPARQL. Currently, to obtain inference results of given SPARQL queries, Web ontology repositories construct inference ontology models and match the SPARQL queries with the models. However, an inference model requires much larger space than its original base model, and reusability of the model is not available for other inferrable SPARQL queries. Therefore, the aforementioned approach is not suitable for large scale SPARQL query processing. To resolve tills issue, this paper proposes a novel SPARQL query rewriting algorithm that can obtain results by rewriting SPARQL queries and accomplishing query operations against the base ontology model. To achieve this goal, we first define OWL-DL inference rules and apply them on rewriting graph pattern in queries. The paper categorizes the inference rules and discusses on how these rules affect the query rewriting. To show the advantages of our proposal, a prototype system based on lena is implemented. For comparative evaluation, we conduct an experiment with a set of test queries and compare of our proposal with the previous approach. The evaluation result showed the proposed algorithm supports an improved performance in efficiency of the inferrable SPARQL query processing without loss of completeness and soundness.

• Journal of Korea Spatial Information System Society
• /
• v.12 no.1
• /
• pp.18-27
• /
• 2010

In u-GIS environment, GeoSensor environment requires that dynamic data captured from various sensors and static information in terms of features in 2D or 3D are fused together. GeoSensors, the core of this environment, are distributed over a wide area sporadically, and are collected in any size constantly. As a result, storage space could be exceeded because of restricted memory in DSMS. To solve this kind of problems, a lot of related studies are being researched actively. There are typically 3 different methods - Random Load Shedding, Semantic Load Shedding, and Sampling. Random Load Shedding chooses and deletes data in random. Semantic Load Shedding prioritizes data, then deletes it first which has lower priority. Sampling uses statistical operation, computes sampling rate, and sheds load. However, they are not high accuracy because traditional ones do not consider spatial characteristics. In this paper 'Pre-Filtering based Post Load Shedding' are suggested to improve the accuracy of spatial query and to restrict load shedding in DSMS. This method, at first, limits unnecessarily increased loads in stream queue with 'Pre-Filtering'. And then, it processes 'Post-Load Shedding', considering data and spatial status to guarantee the accuracy of result. The suggested method effectively reduces the number of the performance of load shedding, and improves the accuracy of spatial query.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1992.10a
• /
• pp.3-3
• /
• 1992

타이어 Curing공정은 공기압 타이어의 제조시 상당히 정교한 단계를 거쳐서 이루어지며, 이는 타이어 설계에 큰 영향을 줄 뿐만아니라, 타이어의 성능에도 관건이 있다. 본 연구에서는 유한요소법을 이용하여 타이어의 molding 공정을 분석하였다. 유한요소해석 프로그램인 MARC가 Cured 타이어 내부의 Curing Bladder 팽창과정해석에 이용되었다. 비압축성 요소로 Curing Bladder를 모형화하였으며, MARC의 접촉문제해석기법(contact option)을 이용하여 Cured 타이어 내부와 Curing Bladder 외부의 접촉부위를 Simulation하였다. 본 연구의 주요 관심내용으로서는 Curing Bladder의 형상변화에 따른 Curing Bladder의 팽창거동해석과, Cured타이어와 Curing Bladder의 접촉부위에서 얻을 수 있는 접촉압력의 비교.검토이다. 타이어 Curing시 타이어와 Bladder의 Contact과정을 해석하여, 아래와 같은 결과를 도출하였다. Bladder의 형상은 Cylinderical 형상 보다는 Toroidal 형태가 접촉압 분포의 균일성 및 크기 측면에 서 우수한 것으로 판단된다. Curing Bladder의 증심선 부위 보다 이에서 약간 떨어진 부위에서 최대 접촉압력이 발생되며, 이는 타이어 내면의 굴곡현상과 깊은 관련이 있윰 것으로 사료된다. 타이어 Bead부의 Carcass 자연평형현상이 유지된 제품을 얻기위해서는, Side-Bead구간의 접촉압력 증가가 필요하며, 이를 위하여는 Bladder 형상이 Cylinderical 보다는 Toroidal 형태가 유리하고, Bead부의 Gage Down, 전체직경의 증가 및 높이의 증가가 유리한 것으로 판단된다. 본 연구 결과를 이용하여, 타이어 Curing과정에서 발생되는 불량제품의 원인파악 및 타이어 설계자가 원하는 제품생산의 불가능한 원인을 파악하는데 도움을 줄 것이다.를 C의 structure와 pointer를 기반으로 하게끔 변경시키고 이에 따르는 제반 변경 사항을 수정 보완하여 프로그램의 분석을 용이하게 하며 기능의 변경 및 추가가 수월하게 하였고 메모리를 동적으로 관리할 수 있게 하였다. 또한 기존의 smpl에 디버깅용 함수 및 설비(facility) 제어용 함수를 추가하여 시뮬레이션 프로그램 작성을 용이하게 하였다. 예를 들면 who_server(), who_queue(), pop_Q(), push_Q(), pop_server(), push_server(), we(), wf(), printfct() 같은 함수들이다. 또한 동시에 발생되는 사건들의 순서를 조종하기 위해, 동시에 발생할 수 있는 각각의 사건에 우선순위를 두어 이 우선 순위에 의하여 사건 리스트(event list)에서 자동적으로 사건들의 순서가 결정되도록 확장하였으며, 설비 제어방식에 있어서도 FIFO, LIFO, 우선 순위 방식등을 선택할 수 있도록 확장하였다. SIMPLE는 자료구조 및 프로그램이 공개되어 있으므로 프로그래머가 원하는 기능을 쉽게 추가할 수 있는 장점도 있다. 아울러 SMPLE에서 새로이 추가된 자료구조와 함수 및 설비제어 방식등을 활용하여 실제 중형급 시스템에 대한 시뮬레이션 구현과 시스템 분석의 예를 보인다._3$", chain segment, with the activation energy of carriers from the shallow trap with 0.4[eV], in he amorphous regions.의 증발산율은 우기의 기상자료를 이용하여 구한 결과 0.05 - 0.10 mm/hr 의 범위로서 이로 인한 강우손실량은 큰 의미가 없음을 알았다.재발이 나타난 3례의 환자를 제외한 9례 (75%)에서는 현재까지 재발소견을 보이지 않고 있다. 이러한 결과는 다른 보고자들과 유사한 결과를 보이고 있지만 아직까지 증례가 많지 않기 때문에 생존율을 얻

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.5
• /
• pp.169-176
• /
• 2020

Recently, deep neural networks (DNNs) have brought revolutions to many mobile and embedded devices by providing human-level machine intelligence for various applications. However, high inference accuracy of such DNNs comes at high computational costs, and, hence, there have been significant efforts to reduce computational overheads of DNNs either by compressing off-the-shelf models or by designing a new small footprint DNN architecture tailored to resource constrained devices. One notable recent paradigm in designing small footprint DNN models is sharing parameters in several layers. However, in previous approaches, the parameter-sharing techniques have been applied to large deep networks, such as ResNet, that are known to have high redundancy. In this paper, we propose a parameter-sharing method for already parameter-efficient small networks such as ShuffleNetV2. In our approach, small templates are combined with small layer-specific parameters to generate weights. Our experiment results on ImageNet and CIFAR100 datasets show that our approach can reduce the size of parameters by 15%-35% of ShuffleNetV2 while achieving smaller drops in accuracies compared to previous parameter-sharing and pruning approaches. We further show that the proposed approach is efficient in terms of latency and energy consumption on modern embedded devices.