• Journal of Korea Society of Industrial Information Systems
• /
• v.5 no.2
• /
• pp.22-31
• /
• 2000

The VB5.2 interface of B-lSDN, that is located between an access network and a service node, allows dynamic allocation and release of ATM resources. In this paper, we propose the B-ANCC protocol of the VB5.2 interface to minimize the overall connection setup delay by introducing the access network. The B-ANCC protocol enhances the B-BCC protocol and adopts a parallel interworking function with signaling protocols in the service node. To confirm the correctness of the proposed B-ANCC protocol, we validate it using the automated validation tool, SPIN. We analyze and simulate the sequential interworking model based on the B-BCC protocol and the parallel interworking model based on the B-ANCC protocol, in terms of a connection setup delay and a completion ratio. It is shown that our proposed parallel interworking model with B-ANCC reduces a setup delay and improves a completion ratio compared to the sequential interworking model with B-BCC.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.05a
• /
• pp.48-51
• /
• 2019

최근 기술이 발달함으로 인해 더 짧은시간에 더 많은 계산량이 필요해진 시대가 왔다. 본 연구에서는 CPU와 GPU의 구조를 파악하고 계산속도를 비교한다. 직렬 방식의 알고리즘에서의 병렬 방식의 알고리즘 및 현재 GPU 병렬처리 적용 사례 및 추후 적합한 모델 찾기에 대해 연구한다.

• The KIPS Transactions:PartB
• /
• v.14B no.3 s.113
• /
• pp.231-240
• /
• 2007

In this paper, we propose an automatic extraction model for unknown translations and implement an unknown translation extraction system using the proposed model. The proposed model as a phrase-alignment model is incorporated with three models: a phrase-boundary model, a language model, and a translation model. Using the proposed model we implement the system for extracting unknown translations, which consists of three parts: construction of parallel corpora, alignment of Korean and English words, extraction of unknown translations. To evaluate the performance of the proposed system we have established the reference corpus for extracting unknown translation, which comprises of 2,220 parallel sentences including about 1,500 unknown translations. Through several experiments, we have observed that the proposed model is very useful for extracting unknown translations. In the future, researches on objective evaluation and establishment of parallel corpora with good quality should be performed and studies on improving the performance of unknown translation extraction should be kept up.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.01a
• /
• pp.283-284
• /
• 2021

본 논문에서는 서로 다른 진화 전략의 병렬화를 구현하기 위해 섬 모델을 도입하고 자원 간의 균형을 유지하기 위해 Monod 모델을 활용하는 PDE-EM이라는 생태 모델 알고리즘을 기반으로 한 새로운 병렬 DE를 제안하도록 한다. 각 섬은 동일한 자원으로 서로 다른 전략으로 진화한다. 지정된 세대 수마다 섬의 진화 정도에 따라 등급이 매겨지고, Monod 모델을 활용하여 각 섬에 다양한 자원이 할당된다.

• The KIPS Transactions:PartA
• /
• v.10A no.1
• /
• pp.25-34
• /
• 2003

Parallel file system solves I/O bottleneck to store a file distributedly and read it parallel exchanging messages among computers that is connected multiple computers with high speed networks. However, they do not consider the message characteristics and performances are decreased. Accordingly, the current study proposes the Hybrid Channel model (HCM) as a message-management method, whereby the messages of a parallel file system are classified by a message characteristic between control messages and file data blocks, and the communication channel is divided into a message channel and data channel. The message channel then transfers the control messages through TCP/IP with reliability, while the data channel that is implemented by Virtual Interface Architecture (VIA) transfers the file data blocks at high speed. In tests, the proposed parallel file system that is implemented by HCM exhibited a considerably improved performance.

• Nuclear Medicine and Molecular Imaging
• /
• v.43 no.5
• /
• pp.443-450
• /
• 2009

Purpose: Conventional image reconstruction uses simplified physical models of projection. However, real physics, for example 3D reconstruction, takes too long time to process all the data in clinic and is unable in a common reconstruction machine because of the large memory for complex physical models. We suggest the realistic distributed memory model of fast-reconstruction using parallel processing on personal computers to enable large-scale technologies. Materials and Methods: The preliminary tests for the possibility on virtual manchines and various performance test on commercial super computer, Tachyon were performed. Expectation maximization algorithm with common 2D projection and realistic 3D line of response were tested. Since the process time was getting slower (max 6 times) after a certain iteration, optimization for compiler was performed to maximize the efficiency of parallelization. Results: Parallel processing of a program on multiple computers was available on Linux with MPICH and NFS. We verified that differences between parallel processed image and single processed image at the same iterations were under the significant digits of floating point number, about 6 bit. Double processors showed good efficiency (1.96 times) of parallel computing. Delay phenomenon was solved by vectorization method using SSE. Conclusion: Through the study, realistic parallel computing system in clinic was established to be able to reconstruct by plenty of memory using the realistic physical models which was impossible to simplify.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.3-8
• /
• 2023

데이터 증강은 인공지능 모델의 학습에서 필요한 데이터의 양이 적거나 편향되어 있는 경우, 이를 보완하여 모델의 성능을 높이는 데 도움이 된다. 이미지와는 달리 자연어의 데이터 증강은 문맥이나 문법적 구조와 같은 특징을 고려해야 하기 때문에, 데이터 증강에 많은 인적자원이 소비된다. 본 연구에서는 복수의 대규모 언어 모델을 사용하여 입력 문장과 제어 조건으로 프롬프트를 구성하는 데 최소한의 인적 자원을 활용한 의미적으로 유사한 문장을 생성하는 방법을 제안한다. 또한, 대규모 언어 모델을 단독으로 사용하는 것만이 아닌 병렬 및 순차적 구조로 구성하여 데이터 증강의 효과를 높이는 방법을 제안한다. 대규모 언어 모델로 생성된 데이터의 유효성을 검증하기 위해 동일한 개수의 원본 훈련 데이터와 증강된 데이터를 한국어 모델인 KcBERT로 다중 클래스 분류를 수행하였을 때의 성능을 비교하였다. 다중 대규모 언어 모델을 사용하여 데이터 증강을 수행하였을 때, 모델의 구조와 관계없이 증강된 데이터는 원본 데이터만을 사용하였을 때보다 높거나 그에 준하는 정확도를 보였다. 병렬 구조의 다중 대규모 언어 모델을 사용하여 400개의 원본 데이터를 증강하였을 때에는, 원본 데이터의 최고 성능인 0.997과 0.017의 성능 차이를 보이며 거의 유사한 학습 효과를 낼 수 있음을 보였다.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.4 s.42
• /
• pp.35-45
• /
• 2006

RIn this paper. we describe real-time parallel processing simulator developed for the use of performance analysis of rolling missiles. The real-time parallel processing simulator developed here consists of seeker emulator generating infrared image signal on aircraft, real-time computer, host computer, system unit, and actual equipments such as auto-pilot processor and seeker processor. Software is developed according to the design requirements of mathematic model, 6 degree-of-freedom module, aerodynamic module which are resided in real-time computer. and graphic user interface program resided in host computer. The real-time computer consists of six TI C-40 processors connected in parallel. The seeker emulator is designed by using analog circuits coupled with mechanical equipments. The system unit provides interface function to match impedance between the components and processes very small electrical signals. Also real launch unit of missiles is interfaced to simulator through system unit. In order to use the real-time parallel processing simulator developed here as a performance analysis equipment for rolling missiles, we perform verification test through experimental results in the field.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.42 no.5 s.335
• /
• pp.1-8
• /
• 2005

A RC parallel gate model is used to consider gate distributed effect that affects RF MOSFET performance, and extraction formula based on $Y_{11}$-parameters are used to extract model parameters directly from measured S-parameters. Better agreement between measured and modeled S-parameters in the frequency range beyond 10 GHz is achieved by using the RC parallel model than conventional Rg one, demonstrating the accuracy of the RC model and extraction technique. Using these extraction methods, gate voltage dependent curves of RC gate model parameters are newly extracted, and these parameter data will greatly contribute to developing a RF nonlinear gate model.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04a
• /
• pp.615-617
• /
• 2000

인터넷 상에 분산되어 있는 다수의 일반 컴퓨터들을 이용하여 높은 컴퓨팅 파워를 요구하는 응용문제를 병렬분산 처리함으로써 값 비싼 고성능의 수퍼컴퓨터를 사용하는 것 보다 경제적인 효과를 얻을 수 있다. 본 연구에서는 다중 에이전트 시스템을 이용해서 가장 병렬 시스템을 구성하고, 기존 방법들 보다 더 효과적인 방법으로 워커 에이전트와 작업 패키지를 분산하고 결과를 얻는 새로운 방법을 제안한다. IBM의 Aglet 시스템을 이용하여 이동 컴퓨팅 환경을 모델링 하였고, 제안된 분산 기법에 관한 성능 모델을 수학적으로 유도하여 그 결과를 기존 결과와 비교함으로써 본 논문에서 제안된 방법이 더 효율적임을 보인다.