• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04b
• /
• pp.529-531
• /
• 2002

생방송 비디오 프록시는 사용자에게 최신의 비디오 데이터를 전송해야 한다. 그러나 큰 파일을 나누어 미리 캐싱하는 기존의 방법으로는 최신의 데이터를 보여줘야 하는 생방송 VOD의 제약을 만족시킬 수 없다. 본 논문에서는 생방송 비디오 스트림을 대상으로 효과적으로 캐싱할 수 있도록 클라이언트의 요청을 받는 큐와, 생방송 비디오 데이터를 저장하는 버퍼풀을 구성하였고, 이들이 제대로 동작하기 위한 동기화 과정을 설계하였다. 이 구조의 특징은 프록시가 클라이언트들의 요청을 일정시간 동안 모아서, 미디어 서버에게 한번만 요청한다는 데에 있다 미디어 서버는 프록시에게만 요청을 받고 데이터를 전달하게 되며, 모든 부하는 프록시 쪽으로 옮겨간다 결국 프록시 서버의 개수를 늘리거나 성능을 확장하면 서비스 가능한 클라이언트의 수를 확장시진 수 있는, 이른바 확장성의 개선을 기대할 수 있다. 실제로 구현해본 절과 지연시간 개선, 확장성 효과 등 프록시로서 만족할 만한 성능이 나옴을 확인하였다. 비연속적인 비디오 데이터를 대상으로 한 본 논문은, 오디오와 같이 연속적인 성질의 데이터에 대한 처리를 보강한다면 수요가 늘고있는 생방송 VOD서비스에 대비한 프록시 개발에 도움을 줄 것이다.

• 한국IT서비스학회:학술대회논문집
• /
• 2003.05a
• /
• pp.607-614
• /
• 2003

멀티미디어 스트림 전송에 있어서 대량접속자를 수용하기 위해서는 서버에 많은 리소스와 대역폭이 집중되어야만 하는데, 소수의 클라이언트 접속만을 수용할 수 있고 적은 리소스와 대역폭을 갖는 서버의 경우, 이미 접속한 클라이언트가 서버 기능을 해서 중계할 수 있다면, 추가적인 접속은 중계가 가능한 다른 클라이언트로 접속하게 할 수 있다. 이렇게 피라미드 형태로 접속이 분산하기 위해서 P2P 접속을 이용한다면, 적은 리소스와 대역폭을 가지고도 대량접속자를 수용할 수 있는 대안이 될 수 있다. 실제로 이러한 구조로 어떤 어플리케이션이 가능한지, 또 얼만큼의 성능개선효과가 있는지 고찰해보고자 한다.

• The Journal of Society for e-Business Studies
• /
• v.18 no.3
• /
• pp.159-175
• /
• 2013

Recently, as real-time availability of e-commerce data becomes possible, the requirement of real-time analysis of e-commerce increases significantly. In the real-time analysis of e-commerce data, it is very important to efficiently process continuous join queries between an e-commerce data stream and disk-based large relations. In this paper, we propose an efficient method for processing a continuous join query between an e-commerce data stream and multiple disk-based relations. The proposed method improves the service rate significantly, while reducing the amount of required memory substantially. Through analysis and various experiments, we show the efficiency of the proposed method compared with the previous one in terms of service rate and memory usage.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.11C
• /
• pp.1112-1123
• /
• 2002

Providing a multimedia service requires efficient network bandwidth allocation scheduling. Typically, transmission of information with bursty characteristics, such as a video data stream, utilizes some type of smoothing technique that reduces the transmission rate fluctuation. While capable of a drastic reduction in the peak transmission rate, even the optimum smoothing algorithm proposed to date, however, is rather inadequate for multi-stream applications that share network bandwidth since such an algorithm has been designed and applied toward handling each independent stream. In this paper, we proposed an efficient bandwidth allocation algorithm that addresses these shortcomings. The proposed bandwidth allocation algorithm is divided into two steps, a static bandwidth allocation and a dynamic bandwidth allocation. In the former case, the peak rate reduction is achieved by maintaining the accumulated data in the user buffer at a high level, whereas the concept of real demand factor is employed to meet the needs of multi-stream bandwidth allocation in the latter case, Finally, the proposed algorithm's performance was verified with ExSpect 6.41, a Petri net simulation tool.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.4
• /
• pp.521-528
• /
• 2016

Web-based services are used widely in many computer application fields due to the increasing use of PCs and mobile devices. Accordingly, topics on the analysis of access logs generated in the application fields have been researched actively to support personalized services in the field, and analyzing techniques based on the weight differentiation of information in access logs have been proposed. This paper outlines an analysis technique for web-click streams, which is useful for finding high utility old item sets in web-click streams, whose data elements are generated at a rapid rate. Using the technique, interesting information can be found, which is difficult to find in conventional techniques for analyzing web-click streams and is used effectively in target marketing. The proposed technique can be adapted widely to analyzing the data generated in a range of computing application fields, such as IoT environments, bio-informatics, etc., which generated data as a form of data streams.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.2
• /
• pp.79-85
• /
• 2008

The middleware that provides RFID-based ubiquitous application service should process the data inputted constantly in real time, and acquire and deliver the answers of the questions in the application service. Studies for developing a Data Stream Management System(DSMS) has been performed in order to process a large amount of data stream inputted constantly in this way. Previous algorithms on data stream were mostly focused on reducing the average error between the answers of the successive questions and abandon the data according to the priority of them when a load occurs. This article is composed of presenting the necessity of the studies on the DSMS and speedy data processing, suggesting an algorithm to make Possible the speedy data processing using buffers and prompt questions and answers, and testing the performance of the data processing rate and whether a buffer is generated correspondingly to the algorithm suggested, in either a single or a multiple buffer, through simulations.

• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.9
• /
• pp.527-535
• /
• 2004

Handling mixed workload in digital set-top box or streaming server becomes an important issue as integrated file system gets momentum as the choice for the next generation file system. The next generation file system is required to handle real-time audio/video playback while being able to handle text requests such as web page, image file, etc. Legacy file system provides only best effort I/O service and thus cannot properly support the QoS of soft real-time I/O. In this paper, we would like to present our experience in developing the file system which fan guarantee the QoS of multimedia stream. We classify all application I/O requests into two category: periodic I/O and sporadic I/O. The QoS requirement of multimedia stream could be guaranteed by giving a higher priority to periodic requests than sporadic requests. The proto-type file system(Qosfs) is developed on Linux Operating System.

• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.8
• /
• pp.445-452
• /
• 2004

Zoning technology of disk improved the performance of disk subsystem with increase of storage capacity and average transfer bandwidth. SCAN disk scheduling with double buffering is used to utilize the performance of zoned disk in multimedia system. However, this method has a problem that generates jitter when the number of steams increases. In this article, we propose the novel approach, pre-buffering policy, to overcome this problem. Pre-buffering avoids jitter by buffering the lack of data before starting service, which is estimated from the current cycle length and the maximum cycle length. We can calculate cycle length, data sire needed in each cycle and the possible lack of data caused by the increase of the number of streams using the numerical model of disk subsystem. Pre-buffering can be applied for multimedia systems and contribute to provide clients with high quality service without jitter.

• Journal of Korea Multimedia Society
• /
• v.2 no.3
• /
• pp.277-288
• /
• 1999

A distributed multimedia school system is developed for the multimedia classroom at high school and university. The system is designed and implemented for students to improve the learning efficiency through the personalized multimedia contents and pace of learning. The previously developed multimedia information retrieval systems have some limitations on being applied to the multimedia classroom: expensive cost per stream or poor retrieval quality inappropriate for education, unscalability of system and service, unfamiliar proprietary client environment, and difficulty for teachers to use the authoring tools and manage the authored teaching materials. The system we developed overcomes the above problems. It is so scalable as to be applicable not only to a segmented classroom but also to the world wide Internet. The stream server is one of the components of the system: stream servers clients, a service gateway system, and a authoring management system. This paper describes the design and implementation of the stream server. A single stream server can simultaneously playback the multimedia streams as many as clients at one classroom. This is achieved only by the software engine without any changes of the hardware architecture. The systematic coupling with other components gives the scalability of the system and the flexibility of services.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.13 no.6
• /
• pp.2713-2719
• /
• 2012

According to increasing of various location-based services, distributed data stream processing techniques have been widely studied to provide high scalability and availability. In previous researches, in order to balance the load of distributed nodes, the geographic characteristics of spatial data stream are not considered. For this reason, distributed operations for adjacent spatial regions increases the overall system load. We propose a operation allocation scheme considering the characteristics of spatial operations to effectively processing spatial data stream in distributed computing environments. The proposed method presents the efficient share maximizing approach that preferentially distributes spatial operations sharing the common query regions to the same node in order to separate the adjacent spatial operations on overlapped regions.