XML 기반의 대화형 디지털 TV 재생기 구현

임충수 충남대학교 대학원 2002 국내석사

네트워크 혼잡도 인지 기반의 동적 적응 콘텐츠 스트리밍 기법

민옥기 忠南大學校 大學院 2010 국내박사

With issues related to the 3-screen service and smart phone recently, the adaptive content streaming schemes on the internet environments has been more interesting to many streaming service vendors. Static adaptive content streaming is effective to reflect client device’s capacity and user’s preference, but has a limitation in the point of that it couldn’t adapt to a real-time changeable environment such as internet congestion or battery status of terminal. Existing three kinds of adaptation methods-client based adaptation, server based adaptation, and hybrid adaptation method-, are also insufficient to reflect adaptation depend on the degree of network congestion. In this paper, first of all we presented a parameter reflecting well the degree of network congestion. And then we proposed the scalable video adaptation scheme and system that recognizing degree of network congestion depend on the parameter. The parameter recognizing the degree of network congestion has adopted media RTT. Dynamic adaptation algorithm that proposes in this paper has as follows. First, it selects an initial scalable layer considering the client’s capacity and user’s preference. Second, it changes the scalable layer based on the congestion status dynamically. Lastly, it adjusts the layer again if there is a data packet loss or packet delay but no decreasing with the congestion status. In order to demonstrate the superiority of the dynamic adaptation algorithm based on RTT, we took experiment on QoS elements that give the largest effect to disconnection. Disconnection is one of sensitive elements to client who using contents. In this paper, We experimented and compared proposed the dynamic adaptation algorithm based on RTT with conventional methods such as dynamic adaptation mechanism based on client, dynamic adaptation mechanism based on server, and hybrid dynamic adaptation mechanism. Experimental conditions were established basic contents (MPEG-2 3Mbps, 5min, 30 frames/sec) with range of scalable level 4. Then we have conducted experiments basic contents using stream transfer that generate equivalent complexity. The experimental results show that proposed dynamic adaptation mechanism based on RTT was measured 1.43 higher than dynamic adaptation mechanism based on server. The mean frame rate is increased up to 14.3% compared to the hybrid based dynamic adaptation scheme. The experimental results indicate that average scalable level and frame rate are higher than legacy mechanisms, and packet loss rate is lower compared to legacy mechanisms. We also described dynamic adaptation contents streaming system using mechanism recognizing degree of network congestion proposed in this paper. We made the dynamic adaptive content streaming system higher adaptiveness using the RTT based dynamic adaptation algorithm. And we proposed an expression structure of the scalable layers in MPEG-21 DI meta-data format. For the commercial service, we also described a dynamic streaming service management method. It could get cost effectiveness because of the power saving.

상황인지기반의 고수준 인터페이스를 지원하는 USN 미들웨어

임문섭 忠南大學校 大學院 2010 국내석사

USN이 확장되어 활용되는 분야가 넓어짐에 따라 USN 응용프로그램이 복잡해지고 미들웨어에 대한 요구사항이 많아지게 되었다. 응용프로그램의 개발자의 입장에서 볼 때, USN의 활용 범위가 넓을수록 응용프로그램의 요구사항이 많아질수록 응용프로그램의 개발 또한 복잡해져 개발의 용이성이 낮을 수밖에 없다. 또한 기존의 미들웨어는 응용프로그램과의 통신을 위해 공용인터페이스를 제공하지만 이러한 인터페이스를 통해 얻을 수 있는 정보는 센서가 센싱한 결과인 센서 데이터와 같은 낮은 레벨의 데이터만을 가져올 수 있어 응용프로그램은 데이터를 직접 가공하여 원하는 결과를 만들어내는 과정이 필요하게 된다. 이러한 점은 응용 개발의 난이도를 높여 센서 네트워크의 활용을 어렵게 할 수 있다. 본 논문에서는 상황인지 기술을 이용한 상위레벨 인터페이스를 제공하여 기존의 단순한 정보를 응용에 전달하는 인터페이스의 기능에서 벗어나 상위레벨의 정보를 응용프로그램에 제공하게 함으로써 개발의 편의성을 높이고 응용프로그램에서의 연산 부하를 감소시키는 것을 목적으로 한다. 더불어 센서 네트워크의 미들웨어에서 반드시 지원해야 하는 센서 네트워크 추상화를 위해 SensorML를 사용하여 다른 센서 네트워크를 위한 미들웨어와의 정보 통합의 용이함은 물론 표현의 차이로 인한 문맥의 차이를 줄임으로써 일관성을 높일 것이다. According to expanding USN(Ubiquitous Sensor Network), USN applications are getting complicated and USN middlewares have to support more requirements than before. In application developer's view, it's getting difficult to develop USN applications because of a lot of requirements. Furthermore existing middlewares provide common interfaces for communication with applications but the interfaces can support only low level sensor data which is made by sensors. It means that low level data processing is needed on USN applications. In this work, the interface of the middleware provides high level data as context-aware results instead of low level data as sensor data through simple interfaces. It means that it will be easier to develop USN applications and the load to applications is less. Moreover this middleware support a metadata based on SensorML[1] which is an information model language that describe either single sensor or sensor platform. In this work, the middleware uses the metadata to be reduce the lack of semantics for sensors and to be easier to integrate the informations between USN middlewares.

정보영재의 프로그래밍 개념 학습을 위한 스크래치 활용 교육 수업 설계

채지희 忠南大學校 敎育大學院 2010 국내석사

Gifted Students in informatics think out various ways of problem solving through programing, and through the process of choosing suitable methods among them, they also acquire higher cognitive ability, such as logical thinking ability and problem solving ability. Besides, they can check out their thinking process by resolving problems through error correction process. However, the curriculums of the programing education for the gifted students in informatics, which are carried out currently in many science education institutes for the gifted, donot match the stages of their cognitive development and may disturb the acquisition of high-level functions, such as improvement of high cognitive function, since they focus on the way of using languages and grammars. Accordingly, a lesson design which could induce learners' interest and inner motivation is necessary for the practice of effective programing concept education for the gifted students in informatics. This study designed a programing class for the gifted students in informatics by using educational programing language, Scratch, and by grafting Renzulli's Enrichment Triad Model onto Treffinger's Self-directed Learning Model among teaching and learning models for the gifted children. The use of Scratch in programing education for the gifted students in informatics offers newer and more interesting programing learning experiences than the conventional programming languages, like Visual Basic and C, and is yet similarly effective in concept acquisition. Besides, it has a possibility of improving high cognitive ability, such as creativity and problem solving ability, through the process of completing programing works, such as animations and games, and through the sharing of the programs. 정보 분야의 영재들은 프로그래밍을 통해 문제 해결을 위한 다양한 방법을 생각해 내고 그 중에서 적합한 방법을 선택하는 과정을 통해 논리적 사고력, 문제분석능력과 같은 고등인지기능을 습득할 수 있다. 또한 오류 수정 과정을 통하여 스스로 문제를 해결함으로써 자신의 사고 과정을 점검할 수 있다. 그러나 현재 대학부설 과학영재교육원 등의 정보 분야 교육에서 실시되고 있는 프로그래밍 교육은 영재들을 대상으로 한다는 이유로 학습자의 인지발달단계에 맞지 않는 교육과정으로 진행되고 있으며, 프로그래밍을 교육함에 있어서 언어의 사용법 및 문법에 중점을 두기 때문에 고등인지 기능의 신장과 같은 고차원적인 기능의 습득을 저해 할 수 있다. 따라서, 정보 분야의 영재들에게 효과적인 프로그래밍 개념 교육을 실시하기 위해서는 학습자의 흥미와 내적동기를 부여할 수 있는 수업 설계가 필요하다. 본 연구에서는 MIT에서 개발한 교육용 프로그래밍 언어인 스크래치를 활용하고, 영재 교수-학습 모형 중 Renzulli의 심화학습 3단계 모형과 Treffinger의 자기주도적 학습모형을 접목하여, 정보 분야의 영재들을 위한 스크래치 활용 프로그래밍 교육의 수업을 설계하였다. 이렇게 설계된 수업의 일부분을 충남대학교 과학영재교육원 정보 분야 기초과정(중학교 1학년) 학생들에게 실험 적용해봄으로써, 스크래치를 활용한 정보영재의 프로그래밍 교육이 Visual Basic, C와 같은 전통적인 언어보다 새롭고 흥미로운 프로그래밍 경험을 제공하면서도 프로그래밍 개념 습득에는 동일한 효과를 낼 수 있으며, 또한 애니메이션, 게임 등의 완성된 작품을 스스로 프로그래밍 하는 과정과 프로그램의 공유를 통해 창의력, 문제해결력과 같은 고등인지 능력을 신장시킬 수 있는 가능성이 있음을 알 수 있었다.

원자력발전소 주제어실의 프로액티브 경보감축방법 및 인간공학적 유효성검증

장귀숙 忠南大學校 大學院 2010 국내박사

The nuclear regulatory agency of KOREA recommended that an alarm system of a main control room (MCR) for nuclear power plants (NPP) should be designed so that inadequate alarms could be eliminated and less important alarms be suppressed operators to be able to focus on important alarms. And the nuclear regulatory agency of KOREA recommended that the effectiveness of an alarm reduction method should be proven through an ergonomics validation test under dynamic simulation environment of a NPP. The existing implementations of various types of alarm reduction systems did not negatively affect the effectiveness of the alarm reduction and did not show an effect on the alarm reduction. The results of existing researches had difficulty in verification and validation of the software because these used an inference engine of an expert system. Also these had difficulty in the feedback of operation experiences for the alarm reduction processing. So, the results of existing researches did not actually apply to a NPP. To overcome these problems of results of existing researches, this thesis proposed a proactive alarm reduction method based on the contents of the past operating affects during the next time of the alarm reduction. Also, this thesis proposed an ergonomic validation test method for the alarm reduction method of a MCR for a NPP. The proactive alarm reduction method obtained ECA (Event Condition Action) rules and facts for alarm reductions based on the operating experiences of operators and verified rules and facts by using the independent reviews of different operators. And the proactive alarm reduction method made that behaviors of ECA rules were deterministic and eliminated and suppressed alarms based on the ECA rules and facts for an alarm reduction under the alarm flood situation. Also, the proactive alarm reduction method provided the alarm log analysis function of the alarm system which produced the verification of facts on the alarm reduction and the performance monitoring data based on alarm log data and fed back the complementary and modifications of rules and facts for alarm reductions. Finally, the proactive alarm reduction method improved the efficiency of alarm reduction. And this thesis proposed an ergonomic validation method of the alarm reduction method. An ergonomic validation test method defined the validation measures that were a cognitive performance time, a cognitive accuracy, an objective situation awareness and a subjective situation awareness which were measured by a situation awareness validation system. And an ergonomic validation test method defined the test environment and a test procedure for ergonomics validation test and analysis methods of test results based on statistics and signal detection theory. And this thesis designed and implemented the proactive alarm reduction system, tested preliminary functional testing, constructed the environment for the ergonomic validation test. Also 8 subjects actual working on a NPP tested the practical effectiveness of the proposed proactive alarm reduction method according to the proposed the ergonomic validation test method under a full scope dynamic simulator of a NPP. The results of the test were that four measurements of proposed alarm reduction method are equal or greater than the existing methods under 53% alarm reduction situation of a MCR. Finally, the proactive alarm reduction method meets the acceptance criteria for applying new MMI (Man-Machine Interface) techniques in a MCR of a NPP. Therefore, this proactive alarm reduction method is expected to apply to new NPP. And this proactive alarm reduction method and this ergonomic validation test method are expected to be inputted to the diagnosis system of a NPP and human factor verification and validation tests of a MCR of a NPP.

이동 객체 정보의 효율적인 저장을 위한 필터링 기법

홍순필 忠南大學校 大學院 2010 국내석사

데이터 정합성을 보장하는 멀티 에이전트 기반 정보상호운용 프레임워크

최기석 忠南大學校 大學院 2013 국내박사

현대의 행정기관 및 공공기관의 업무는 주로 정보시스템을 이용하여 수행되고 있으며, 많은 곳에서 계속적인 정보화 사업 진행으로 새로운 정보시스템들이 구축되어지고 있다. 정보시스템은 업무 고유의 목적에 따라 구축되어 활발히 이용되고 있으며, 많은 서비스를 제공하고 있다. 이러한 공공부문의 업무를 정보화 기술을 이용한 업무 전자화가 이루어진 정보시스템을 바탕으로 효율적으로 수행하는 정부를 ‘전자정부’라 일컫는다. 전자정부의 많은 정보시스템들은 업무와 정책에 따라 구축되고 독립적으로 운영된다. 그러나 하나의 업무를 다수의 정보시스템을 이용하거나 다양한 업무가 하나의 정보시스템에서 운영되는 경우도 발생하며, 특정 업무와 정보에 대하여 효율적인 관리와 처리를 하기 위한 시스템 상호운용과 정보의 공동활용 등에 대한 필요성이 커지고 있다. 본 논문에서는 분산되고 독립적으로 운영되는 정보 시스템들 간의 상호운용을 위한 협력 환경의 인프라 구축 정보연계 프레임워크에서 데이터 신뢰성을 보장하기 위한 방안을 제안한다. 이는 전자정부환경에서 분산 정보시스템들간의 이질정보에 대한 데이터 상호운용을 위한 프레임워크로서 상호운용하고자 하는 데이터의 신뢰성 보장과 가용성 향상을 위하여 데이터 품질관리 방법을 제안하며 전자정부 정보시스템의 보안 처리를 위하여 상호운용 되는 데이터의 보안 처리 방법 등을 제안한다. 이 시스템의 특징을 간략히 살펴보면 데이터의 추출 및 적재 각 단계에서 품질 측정이 가능하며 데이터의 완전성, 유효성, 유일성 일관성에 대한 관리가 가능하고 자바(JAVA)기반 GUI 구현으로 다양한 플랫폼에 적용이 가능하며 멀티에이전트 기반으로 운영되어 연계하고자 하는 소스(source)와 타겟(target)을 다양한 방법으로 구현할 수 있어 확장성이 용이하게 된다. 무엇보다 일반적 ETL에서 제공하지 못하는 분산데이터 연계 시 나타나는 여러 종류의 데이터 이질성에 대해 사용자가 정의한 업무규칙을 바탕으로 해결해 준다. 또한 데이터 값이 일정하지 않고 잦은 변화와 정제가 요구되는 데이터의 양이 많아지는 분산 실시간 정보연계 환경에 대응하기 위한 사용자 정의 규칙을 동적으로 생성하고 관리 할 수 있도록 해준다. A Business of modern government agencies and public institutions have been performed mainly using information systems. And information systems have been built In many places as the agency process information projects continuously. Information systems are built and used actively depending on the purpose of their task-specific goals. They also provide a number of services. It refers 'E-government' that performs administrative work of public sector based on the e-business information systems. A lot of e-government information systems are built and operated independently, depending on the business and policy. But sometimes it happens frequently that one work is operated separately by a number of information systems or a variety of tasks are operated by just one information system. The need for system interoperability and sharing information about the efficient management and processing of specific tasks and information is growing. In this paper, I propose a method to ensure the reliability of data for information linking framework as a infrastructure of collaborative environment which is distributed and independently operated between information systems for their interoperability. This is a framework for data interoperability of the heterogeneous information between distributed information systems in the e-government environment. This framework suggests a data quality management technique for secure the reliability and improve the availability of the data which want to interoperate. Also it proposes a security management of data interoperated for security processing of the e-government information systems. A brief look at the features of this system is as follows. The quality can be measured at each stage of data extraction and loading. And data integrity, validity, uniqueness and consistency management is possible. It can be applied to a wide variety of platforms because it is based on GUI implementation of Java. And it is easily scalable because the source and the target that you want link can be implemented in a variety of ways by a multi-agent based management. Above all, this framework can solve the heterogeneity of different kinds of data, that appear when linking distributed data, based on user-defined business rules which general ETL cannot provide. Also, it allows dynamic creation and management of the user-defined rules. This is to cope with the distributed real-time information associated environment that the data value is not constant and the amount of data require frequent changes and refining increase.

모바일 환경에서 사용자의 상황정보를 고려한 멀티미디어 콘텐츠 자동 관리 시스템

심우제 忠南大學校 大學院 2010 국내석사

By advance of computer hardware, mobile device's performances are improved. As a result, users store and use huge amount of multimedia contents in their mobile device. But, It is difficult and uncomfortable to manage and use these contents cause of a limitation of mobile devices' small display and tiny input interfaces. So, users push a wrong input button or take a long time for looking forward to wanted contents and exchange multimedia contents in PC between those in mobile device. In this paper, we propose the system to manage multimedia contents automatically and make recommended multimedia content playlist. For all these things, we use user's situation informations. And we implement the PMCAM & MCPlayer applied above techniques.

사용자의 심리와 상황을 고려한 맞춤형 모바일 멀티미디어 콘텐츠 제공 기법

이용대 忠南大學校 大學院 2009 국내석사

Today’s users have huge amount of multimedia contents and can access it with mobile device everywhere and anytime because of advance of the internet and computer technologies. But, in existing file structure system, users need much time to find multimedia contents that users wanted in many multimedia contents. Specially, users that use mobile device have a difficulty because of mobile device's small display and interface of poor control. So, intelligent management technique for multimedia contents needs to access a number of multimedia contents more effectively. In this paper, we propose the technique to find easily and enjoy multimedia contents in personal computer and to create a personalized list of the contents for mobile device. For providing personalized mobile multimedia contents we use user's situation and psychology. And we implement the PMMCP(Personalized Mobile Multimedia Contents Provider) applied above techniques. The technique we proposed gave users more satisfaction than Top-N technique that recommends user many used contents.

MRDB 시스템을 위한 효율적인 낙관적 동시성제어 기법

윤명훈 忠南大學校 大學院 2010 국내석사