• Journal of IKEEE
• /
• v.7 no.1 s.12
• /
• pp.97-106
• /
• 2003

In ARQ based error control, imperfect error detection leaves error remains on a packet. Aiming for a reduction of error remains in multicopy transmission ARQ system, we propose a rule of requesting a retransmission and deciding a correct copy, (identified as $(m, \;{\sigma})$rule). While the probability of error remains is reduced by the employment of the $(m, \;{\sigma})$ rule at multicopy transmission ARQ, delay and throughput performance may be degraded in comparison with those of conventional single copy transmission ARQ. Thus, we develop an analytical method to evaluate the performance trade-off in multicopy transmission ARQ following the $(m, \;{\sigma})$ rule. From the numerical results obtained by the analytical method, we investigate the effect of channel characteristics on the performance of error remains, packet loss, throughput, and packet delay, and confirm that the adaptability of the $(m, \;{\sigma})$ rule to conform to various QoS requirements with ease.

• Journal of Intelligence and Information Systems
• /
• v.12 no.3
• /
• pp.67-82
• /
• 2006

We propose the HBEM(Hybrid Behavior Evolution Model) composed of rule classification and evolutionary neural network using rule descriptor and link descriptor for evolutionary behavior of virtual robots. In our model, two levels of the knowledge of behaviors were represented. In the upper level, the representation was improved using rule and link descriptors together. And then in the lower level, behavior knowledge was represented in form of bit string and learned adapting their chromosomes by the genetic operators. A virtual robot was composed by the learned chromosome which had the best fitness. The composed virtual robot perceives the surrounding situations and they were classifying the pattern through rules and processing the result in neural network and behaving. To evaluate our proposed model, we developed HBES(Hybrid Behavior Evolution System) and adapted the problem of gathering food of the virtual robots. In the results of testing our system, the learning time was fewer than the evolution neural network of the condition which was same. And then, to evaluate the effect improving the fitness by the rules we respectively measured the fitness adapted or not about the chromosomes where the learning was completed. In the results of evaluating, if the rules were not adapted the fitness was lowered. It showed that our proposed model was better in the learning performance and more regular than the evolutionary neural network in the behavior evolution of the virtual robots.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.7
• /
• pp.656-664
• /
• 2002

We present a statistical analysis of Korean phonological variations using a Grapheme-to-Phoneme (GPT) system. The GTP system used for experiments generates pronunciation variants by applying rules modeling obligatory and optional phonemic changes and allophonic changes. These rules are derived form morphophonological analysis and government standard pronunciation rules. The GTP system is optimized for continuous speech recognition by generating phonetic transcriptions for training and constructing a pronunciation dictionary for recognition. In this paper, we describe Korean phonological variations by analyzing the statistics of phonemic change rule applications for the 60,000 sentences in the Samsung PBS Speech DB. Our results show that the most frequently happening obligatory phonemic variations are in the order of liaison, tensification, aspirationalization, and nasalization of obstruent, and that the most frequently happening optional phonemic variations are in the order of initial consonant h-deletion, insertion of final consonant with the same place of articulation as the next consonants, and deletion of final consonant with the same place of articulation as the next consonant's, These statistics can be used for improving the performance of speech recognition systems.

• Broadcasting and Media Magazine
• /
• v.25 no.1
• /
• pp.28-41
• /
• 2020

인공지능의 고속화를 위한 인공지능용 혹은 딥러닝용 하드웨어 및 소프트웨어 시스템에 대한 수요가 폭발적으로 증가하고 있다. 또한 딥러닝 모델에 따라 다양한 추론 시스템이 끊임없이 연구되고 소개되고 있다. 최근에는 전세계에서 100개가 넘는 회사들에서 인공지능용 추론 칩을 개발하고 있고, 임베디드 시스템에서 데이터센터 솔루션에 이르기까지 다양한 분야를 위한 것들이 존재한다. 이러한 하드웨어의 개발을 위해서 12개 이상의 소프트웨어 프레임 워크 및 라이브러리가 활용되고 있다. 하드웨어와 소프트웨어가 다양한 만큼 이들을 중립적으로 평가하기가 매우 어려운 실정이다. 따라서 업계 표준의 인공지능을 위한 벤치마킹 및 평가기준이 필요한데, 이러한 요구로 인해 MLPerf 추론이 만들어졌다. MLPerf는 30개 이상의 기업과 200개 이상의 머신러닝 연구자 및 실무자들에 의해 운영되고, 전혀 다른 구조를 갖는 시스템을 비교할 수 있는 일관성 있는 규칙과 방법을 제시한다. MLPerf에 의해 제시된 규칙에 의해 2019년도에 처음으로 다양한 인공지능용 추론 하드웨어가 벤치마킹을 수행했다. 여기에는 14개의 회사에서 600개 이상의 추론 결과를 측정하였으며, 30개가 넘는 시스템이 이러한 추론에 사용되었다. 본 원고에서는 MLPerf의 학습과 추론을 중심으로 하여 최근에 개발된 다양한 회사들의 인공지능용 하드웨어, 즉 가속기 들의 성능을 살펴보고자 한다.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.23 no.2
• /
• pp.243-250
• /
• 2013

We have entered into an era of smart software system where the many kinds of embedded software, especially SCADA and Automotive software not only require high reliability and safety but also high-security. Removing software weakness during the software development lifecycle is very important because hackers exploit weaknesses which are source of software vulnerabilities when attacking a system. Therefore the coding rule as like core functions of MISRA-C should expand their coding focus on security. In this paper, we used CERT-C secure coding rules for nuclear-related software being developed to demonstrate high-safety software, and proposed how to remove software weakness during development.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2000.04a
• /
• pp.443-453
• /
• 2000

장차전의 양상은 고도의 과학전, 타격수단의 다양화 등에 따라 고속이체기동전이 t행될 것이며, 이러한 전쟁양상은 정확하고, 보다 빠른 첩보와 정보의 수집 및 분석을 통하여 아군의 신속한 의사결정 및 대응을 요구한다. 이를 위하여 첩보와 정보 수집 및 분석을 자동화하기 위한 전장정보분석 자동화에 관한 연구가 국방과학연구소 주관으로 실시되고 있다. 따라서 이와 연계된 의사결정 자동화에 관한 연구가 필요하게 되었다. 본 연구는 이러한 요구에 부응할 수 있는 전장정보를 활용한 의사결정의 중요한 한 분야인 화력분배를 자동화하기 위한 전문가시스템의 지식베이스모듈에 대한 분석 및 설계에 관한 연구이다.기존에는 화력분야에 대한 아방책 선정까지를 자동화하는 전문가시스템 개발에 대한 연구가 수행되었으나, 본 연구에서는 자동화의 효율성을 높이기 위해서 아방책 선정에서 나아가 아군 화력자산의 파괴율을 고려하고, 지휘관의 의도에 부합하는 아군 화력자산을 배분하는 전문가시스템 개발에 관한 연구를 수행하였다. 본 연구에서는 화력분배 자동화를 위하여 화력분배와 관련된 현행 업무 관련 지식을 획득 및 분석하고 이를 바탕으로 화력자산 분배를 위한 규칙도출 시 개념설계, 상세설계, 알고리즘제시, 규칙추출예시를 하였으며 본 연구결과의 기대효과는 본문을 참고 바란다.

• Computational Structural Engineering
• /
• v.9 no.1
• /
• pp.23-32
• /
• 1996

A rule based system for foundation design of building structures is developed with CLIPS in this study. The types of foundation and the allowable bearing capacity of supporting soil inferred by the rule based system for selection of foundation type, called SOFTEX, are transferred to a structural design program for building foundation. The allowable bearing capacity is calculated with N values of Standard Penetration Test. The foundation types such as independent spread footing, wall footing, combined footing and mat foundation can be inferred by the foundation merge procedure developed in this study. This procedure is based on the analysis data from the super structure and the estimated bearing capacity. By using this integrated system, structural engineers with less experience in foundation design can design the foundation system for the given superstructure and the site condition with relative ease.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.8
• /
• pp.39-45
• /
• 2020

Despite the increasing interest in cyber security in recent years, the emergence of new technologies has led to a shortage of professional personnel to efficiently perform the cyber security. Although various methods such as cyber rage are being used to cultivate cyber security experts, there are problems of limitation of virtual training system, scenario-based practice content development and operation, unit content-oriented development, and lack of consideration of learner level. In this paper, we develop a fuzzy rule-based user-customized training scenario automatic generation system for improving user's ability to respond to infringement. The proposed system creates and provides scenarios based on advanced persistent threats according to fuzzy rules. Thus, the proposed system can improve the trainee's ability to respond to the bed through the generated scenario.

• Journal of Korea Water Resources Association
• /
• v.31 no.6
• /
• pp.641-655
• /
• 1998

The methods of development and application of an expert system are suggested to solve more efficiently the problems of water resources and quality induced by the rapid urbanization. Major parameters of the water quantity and quality of urban areas are selected their characteristics are presented by the sensitivity analysis. The rules to decide the parameters effectively are proposed based on these characteristics. the ESPE(Expert System for Parameter Estimation), an expert system based on the 'facts' and 'rules', is developed using the CLIPS 6.0 and applied to the basin of the An-Yang stream. The results of estimating t도 parameters of water quantity show a high applicability, but those of water quality imply the necessity of improving the present methods due to both the complexity of estimation processes and the lack of decision rules.

• Journal of the Korea Society of Computer and Information
• /
• v.10 no.3 s.35
• /
• pp.161-171
• /
• 2005

SMIL(Synchronized Multimedia Integration Language) allows multimedia objects into a synchronized multimedia presentation. In this paper, I propose modeling rules, formal models, modeling function and an algorithm for SMIL documents using use case diagram and sequence diagram of UML(Unified Modeling Language)to Develop for SMIL Documents Editor. It will be used the use case diagram and collaboration diagram for object-oriented visualizing tool to describe the temporal behavior of the presentation. The main contribution of this paper is that developers of SMIL documents editor easily generate them using this rules and algorithm. And, the formal models and modeling functions provide an environment for processing object-oriented. documents.