http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
ATM망에서 다중 경로 배정 방법을 이용한 고속 대역 예약 기법에 관한 연구
여현 순천대학교 공업기술연구소 1996 工業技術硏究所論文集 Vol.10 No.-
One of the most important services to be supported in ATM networks is high-speed computer communications. However, conventional band-width reservation schemes cannot satisfy the quality of services on the extensive burst traffic and only yield low network efficiency. Bandwidth reservation control methods in ATM network have been the focus of intensive study and various schemes have been proposed recently. In this paper, we investigate burst - level bandwidth reservation schemes with multi - link and multi - path routing and analyze the performance of the protocol. Especially, we analyze two bandwidth reservation modes, DG(Delay Guaranteed) and IU(Immediate Unguaranteed), with multi - link and multi - pah routing respectively. Performance evaluation showed that the delay and blocking probability characteristics are greatly improved, as the number of links between each pair of adjacent nodes, and the number of source - destination paths increases. Propagation delay has a significant impact on the performance. Finally, these schemes improves network efficiency.
패킷 무선망의 프로토롤을 기반으로 한 무선 ATM MAC 프로토콜
여현 순천대학교 기초과학연구소 1998 基礎科學硏究誌 Vol.9 No.-
Most researches on the wireless MAC mainly start from the study on the wireless packet data networks. Several research results are there in the field of wireless packet data networks, but they didn't satisfy the requirements of the Quality of Services for wireless ATM traffic. The reason is that the former wireless networks are designed for each kind of service and so they cannot deal with ATM services which need to serve various kind of services. In order to integrate the various services, common parameters should be defined and efficient structures which support those services and use common parameters should be made. Therefore, in this paper, we propose an efficient MAC structure for wireless ATM and under that structure and algorithm for satisfying the detailed requirements. The results are verified by computer simulations.
다중 패킷 형태의 메세지 입력을 갖는 폴링 시스템에서의 평균 메세지 지연에 관한 연구
여현 순천대학교 공업기술연구소 1993 工業技術硏究所論文集 Vol.7 No.-
The paper presents an analytic method for approximation mean message delays in polling system which is used for modelling computer communication networks. Each message is consisted of multiple packets and the distribution of the number of packts in one message is assumed to be various forms. Using Boxma's pseudo-conservation law, residual cycle time is computed and from residual cycle time mean message delay is derived. Using the result of this study, the transmission delay, under the several conditions of the various traffic types can be approximated well.
로드 밸런싱을 이용한 CBT 멀티캐스트 라우팅 프로토콜의 혼잡제어기법
여현,소순회,이윤주,Yoe, Hyun,So, Sun-Hwe,Lee, Yoon-Ju 한국정보통신학회 2001 한국정보통신학회논문지 Vol.5 No.6
본 연구에서는 내부도메인 멀티캐스트 라우팅 프로토콜중의 하나인 CBT(core-based tree) 프로토콜에서 발생하기 쉬운 코어라우터로의 트래픽 집중 철상을 해결하기 위해 지정코어라우터에 트래픽 폭주가 발생했을 때 멀티캐스트 그룹 가입자로 하여금 동적으로 코어를 변경할 수 있도록 하는 방안을 제시한다. In this paper, we propose a new CBT(core-based tree)multicast routing Mechanism by load balancing mechanism. CBT may result in traffic concentration and bottlenecks near the core routers since traffic from all sources nodes the same set of links as approaches the core. So the router have to load packet to new core router for such congestion. Congestion information is given by rtt(round trip time) between designed core router and the other router.
상호접속된 근거리 통신망 환경에서의 단말간 전송 성능 평가
여현,임재환,최승철,Yoe, Hyun,Lim, Jae-Hwan,Choi, Seung-Chul 한국통신학회 1992 韓國通信學會論文誌 Vol.17 No.4
본 논문에서는 상호 접속된 토큰링 근거리 통신망 환경하에서의 단말간 전송 효율 및 지연 현상에 대한 성능 평가를 행하였다. 상호 접속된 시스템은 단일 및 자중 체인 폐쇄 큐잉 망으로 모델링 하였다. 성능 평가를 위해 평균 값 해석 기법에 기초를 둔 반복 해법 알고리즘을 제안하였다. 단말간 전송 모델의 성능을 평가한 결과 브릿지로 상호 접속한 근거리 통신망 환경하에서 윈도우 크기, 상호 접속 링크 용량, 각 큐의 서비스 시간 등을 망의 구성에 따라 적절히 유지하여야 최적의 전송 효율과 지연을 얻을 수 있음이 발견되었다. In this paper, the performance evaluation on the throughput and the delay between station under the interconnected LANs environments is studied. The connection is formed through bridges. Interconnected system is modelled as single-chain and multiple-chain closed queueing networks. For performance enaluation the iteration algorithm, which is based on the MVA(Mean Value Analysis)is proposed, Various system parameters, such as chain population interconnected communication link capacity, and service times of several queues, are used for the investigation of their influence of the station to station performance, Transport and MAC layers are mainly taken into consideration for the layer modeling. The analysis is shown as a graph by computer programming.
여현,김성진 한국통신학회 2019 정보와 통신 Vol.36 No.3
농업 인구 고령화에 따른 농업 인력의 감소와 산업화에 따른 농업 용지 감소는 우리나라뿐만 아니라 국제적인 문제가 되고 있다. 이로 말미암아 식량 부족 문제가 발생하게 되었고, 이 같은 식량문제 해결을 위해서는 농업 생산성 향상과 노동력 절감이 절실한데 이를 위해 ICT기술을 이용한 스마트팜이 그 해결책으로 제시되고 있다. 국내에서도 스마트팜 분야에 대한 연구 개발과 현장에 대한 확산 보급이 활발히 이루어지고 있다. 하지만 표준화 되지 않은 상태에서 기술개발이 이루어져 시스템 간 비호환성과 AS문제가 대두되고 있다. 이러한 문제를 해결하기 위해 2014년 부터 스마트팜 기술 표준화를 위해 정부와 민간 전문가들이 많은 노력을 기울이고 있다. 본고에서는 국내 스마트팜 표준화 현황과 진행 상황을 보여주고 향후 계획에 대해서 소개한다.
DQDB MAN의 공정성 개선 및 대역폭 향상에 관한 연구
呂鉉 순천대학교 공업기술연구소 1994 工業技術硏究所論文集 Vol.8 No.-
DQDB protocol may not be operated as perfect scheduler when the propagation delay of bus is longer than one slot time. Therefore it results in unfairness problem that the transmission chance of downstream nodes is reduced. So far, for solving this problem, erasure node(EN) scheme has been one of the various solutions, and it is knows for excellent scheme in the bandwidth improvement. But it has the disadvantage of uneasy bandwidth allocation. Particularly, the nodes between erasure nodes show by far the worst throughput. In this paper, the mechanism that EN scheme and proportional assignment (PR)scheme are applied simultaneously is presented, which can improve the throughput of middle nodes and diminish the throughput of head and end nodes. Accordingly, the network bandwidth and fairness are totally improved. After the network model describing the system operation properly is made, its performance is evaluated by simulation. The simulation results show that the fairness has a 25 % increase and the throughput has a 30 % increase compared with the original DQDB protocol. And the throughput has a 9 % decrease but the fairness is improved greatly(80%) compared with using erasure node scheme alone.