• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.11a
• /
• pp.355-356
• /
• 2009

점자 체계와 CMYK 모델을 이용하여 시각 장애인과 일반인이 읽을 수 있는 책을 만들기 위한 방법론을 제시한다. CMYK의 K를 이용해 8개의 점으로 구성된 셀 문자를 텍스트에 삽입하여, 텍스트 스캔 시 빠르고 정확한 문자 인식이 가능하게 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.11a
• /
• pp.387-389
• /
• 2020

오늘날 인스턴트 메시징은 사용자들에게 개인 및 비즈니스 상의 대화의 편리함을 제공하며 거의 모든 현대인들의 일상에 한 축을 담당하고 있다 해도 과언이 아니다. 이러한 모바일 메신저 어플리케이션들은 사용성을 넘어 사용자들의 개인정보 보호와 메시지의 기밀성 및 무결성을 위해 끊임없이 발전해 왔고 많은 사용자들 역시 언론을 통해 발표되는 메신저 어플리케이션들의 여러 보안 이슈들을 접하게 되면서 사용성을 넘어 강력한 보안을 자랑하는 메신저 어플리케이션을 주목하고 이를 사용하고 있다. 그 중 안전하다고 평가받고 있고 백만 이상의 안드로이드 앱 스토어 사용자가 설치한 와이어 라는 메신저 어플리케이션의 프로토콜 분석을 통해 어떤 방법으로 그들이 강조하고 있는 보안 목표를 달성하고 있는지 확인한다.

• Journal of KIISE:Information Networking
• /
• v.32 no.4
• /
• pp.452-461
• /
• 2005

In the star VPN (Virtual Private Network) topology, the traffic between the communicating two CPE(Customer Premise Equipment) VPN GW(Gateway)s nay be inefficiently transferred. Also, the Center VPN GW nav erperience the overload due to excessive packet processing overhead. As a solution to this problem, a direct tunnel can be established between the communicating two CPE VPN GWs using the IKE (Internet Key Exchange) mechanism of IPSec(IP Security). In this case, however, the tunnel establishment and management nay be complicated. In this paper, we propose a mechanism called' SVOT (Star VPN On-demand Tunnel)', which automatically establishes a direct tunnel between the communicating CPE VPN GWs based on demand. In the SVOT scheme, CPE VPN GWs determine whether it will establish a direct tunnel or not depending on the traffic information monitored. CPE VPN GW requests the information that is necessary to establishes a direct tunnel to the Center VPN GW Through a simulation, we investigate the performance of the scheme performs better than the SYST scheme with respect to scalability, traffic efficiency and overhead of Center VPN GW, while it shows similar performance to the FVST with respect to end-to-end delay and throughput.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.9
• /
• pp.87-95
• /
• 2011

Recently, as smartphones have rapidly come into wide use and various social networking services have grown, people perform a variety of interactions through the virtual and real world based on them. Usually, such services focus on easy formation of social links among users, supporting the exchange of simple messages among users on the networks. Twitter, one of such services used worldwide, supports short message service named Tweet and has over 200 million members signed up. n this paper, we propose techniques for supporting real-time group messaging based on the social network of Twitter and describe a smartphone group messaging system developed with the techniques. The system automatically forms a group that include a Twitter user and the followers of the user, supporting real-time group messaging among the group members. The developed system is composed of an XMPP protocol-based messaging server and smartphone client applications which perform real-time messaging based on the social network of Twitter. Twitter users can easily use the system utilizing Tweet messages, exchanging real-time messages with their followers within the groups instantly established on the XMPP server.

• Journal of Korea Game Society
• /
• v.7 no.4
• /
• pp.71-80
• /
• 2007

In this paper, we introduce an efficient GPU-based real-time rendering technique applicable to every kind of game. Our method, without an extra geometry, can represent terrain just with a height map. It makes it possible to freely go around in the air or on the surface, so we can directly apply it to any computer games as well as a virtual reality. Since our method is not based on any geometrical structure, it doesn't need special LOD policy and the precision of geometrical representation and visual quality absolutely depend on the resolution of height map and color map. Moreover, GPU-only technique allows the general CPU to be dedicated to more general work, and as a result, enhances the overall performance of the computer. To date, there have been many researches related to the terrain representation, but most of them rely on CPU or confmed its applications to flight simulation, Improving existing displacement mapping techniques and applying it to our terrain rendering, we completely ruled out the problems, such as cracking, poping etc, which cause in polygon-based techniques, The most important contributions are to efficiently deal with arbitrary LOS(Line Of Sight) and dramatically improve visual quality during walk-through by reconstructing a height field with curved patches. We suggest a simple and useful method for calculating ray-patch intersections. We implemented all these on GPU 100%, and got tens to hundreds of framerates with height maps a variety of resolutions$(256{\times}256\;to\;4096{\times}4096)$.

• Journal of Digital Convergence
• /
• v.12 no.8
• /
• pp.235-242
• /
• 2014

Internet telephony is an Internet service which supports voice telephone using VoIP technology on the IP-based Internet. It has some advantages in that voice telephone services can be accompanied with multimedia services such as video communication and messaging services. In this paper we suggested an Asterisk-based Internet telephony system which can be easily scalable. Most current systems use text files to manage their configuration: SIP users, dialplans, IVR service and etc. But we designed the management system which introduces database tables for efficiency and scalability. It also supports web-based functions developed by using Asterisk, Apache, MySQL, jQuery, PHP and open source softwares.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.12
• /
• pp.2321-2328
• /
• 2006

Today, SIP standard (The signalling protocol for the Internet phone service) raises to be the standard technique because the expandability is high and complexity is low. It is widely investigated and actively advocated to use Si81a1 ring protocol for SIP in VoIP service. SIP service can be applied even outside the Internet phone service; instance messaging and various multimedia technology are just an example. This paper proposed an embodiment proxy server for rambling support to use JAIN SIP API. It provides standard interface for testing the Proxy server for SIP and embodiment of user agent that transfer instant massaging and voice communication.

• Analytical Science and Technology
• /
• v.11 no.4
• /
• pp.235-242
• /
• 1998

Membrane electrode based on chitin(po1y-[$1{\rightarrow}4$]-${\beta}$-N-acetyl-D-glucosamine) was prepared by mixing uniformly grounded of chitin (100 mesh) with PVC and DOS. We investigated the potential response of chitin membrane electrode to metal ions. It was observed that the response slopes for $Cd^{2+}$(34.9 mV/decade) and $Cu^{2+}$(34.0 mV/decade) were larger than those for other ions in pH 4 acetate buffer. The potentiometric response of chitin electrode to varying pH was nearly constant in the pH range of 2~12.

• Journal of KIISE:Computer Systems and Theory
• /
• v.36 no.3
• /
• pp.158-170
• /
• 2009

In this paper, we propose the new torus network which has the hypercube Q3 as the basic module. The proposed Hyper-torus has the degree 4, and is the network which has the scalability, and the fine diameter. If we compare the class of the torus in the viewpoint of network cost, the hyper-torus with $1.4{\sqrt{N}}$+ 16 is proved to be approximately 65% than the torus with $4{\sqrt{N}}$ and 50% than the honeycomb with $2.45{\sqrt{N}}$. This result means that hyper-torus is better for the class of the existing mesh in the viewpoint of network cost.

• Journal of KIISE:Information Networking
• /
• v.28 no.1
• /
• pp.154-163
• /
• 2001

인터넷으로 연결된 다수의 참여자를 지원하는 대규모 분산 가상 환경을 위한 확장성 있는 예측 기반 동시성 제어 방법을 제안한다. 예측 기반 동시성 제어 방식은 낙관적(optimistic) 방식과 같이 참여자들에게 실시간 상호 작용 성능을 제공함과 동시에 비관적(pesimistic)방식과 같이 잠금 허가를 받은 사용자에게만 객체 조작을 허용하므로 충돌을 확실히 방지할 수 있다. 본 논문에서는 사용자 수의 증가에 따른 확장성 있는 예측 알고리즘을 위하여 객체 중심 다중 전송그룹을 도입했다. 객체에 관심있는 객체주변의 사용자들만 객체에 할당된 다중 전송 그룹에 소유권 요청 메시지로서의 참여 메시지를 보냄으로써 소유자 후부가 된다. 현재 소유자는 소유자 후보들 중 다음 소유자를 예측한다. 가상 영역내의 모든 사용자 대신 객체의 할당된 다중 전송 그룹에 소유권 요청 메시지로서의 참여 메시를 보냄으로써 소유자 후보가 된다. 현재 소유자는 소유자 후보들 중 다음 소유자를 예측한다. 가상 영역내의 모든 사용자 대신 객체의 다중 전송 그룹에 참여하고 있는 사용자로부터만 소유권 요청 메시지를 받으므로 소유자가 받는 메시지 수는 가상환경의 전체 사용자의 수에 관계없이 상수값을 갖는다. 이는 소유자의 소유권 요청메세지 처리 시간을 줄여 보다 더 정확한 예측을 하고 사용자의 객체 조작 시간 전에 소유권이 전달되도록한다. 제안한 예측 알고리즘은 라이브러리로 구현되어 기존의 가상 환경 시스템에 적용되었고 실험을 통해 제안한 알고리즘이 대규모 가상 환경에서 갖는 효율성과 확장성을 증명한다.