• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2004.05b
• /
• pp.291-294
• /
• 2004

In this paper, we consider the broadcast switch architecture for hish performance multicast packet switching. In input and output buffered switch, we propose a new switch architecture which supports high throughput in broadcast packet switching with switch planes of single input and multiple output crossbars. The proposed switch architecture has a central arbiter that arbitrates requests from plural input ports and generates multiple grant signals to multiple output ports in a packet transmission slot. It provides high speed pipelined arbitration and large scale switching capacity.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.33 no.5C
• /
• pp.361-370
• /
• 2008

NoC is adopted for data communication between processors and IPs in MPSoC system. NoC has an advantage of scalability in that system can be easily expanded just by adding switches. However, as the number of switches increases, chip area increases as well as data transfer latency. This paper proposes an architecture that can reduce the number of switches in the system by sharing network interfaces. To reduce NI area, the modules sharing network interface use a common buffer in network interface. Experimental results show that the chip area has been reduced by 46.5% and data transfer latency by 17.1%, respectively, compared to conventional architecture.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.4B
• /
• pp.162-169
• /
• 2005

Network traffic measurements show that the data traffic on packet switched networks has the self-similar features which is different from the traditional traffic models such as Poisson distribution or Markovian process model. Most of the call admission control researches have been done on the performance analysis of a single network switch. It is necessary to consider the performance analysis of the proposed admission control scheme under interconnected switch environment because the data traffic transmits through switches in networks. From the simulation results, it is shown that the call admission control scheme may not operate properly on the interconnected switch even though the scheme works well on a single switch. In this parer, we analyze the cell loss probability, utilization and self-similarity of output ports of the interconnected networks switch by using shared buffer memory management schemes and propose the new call admission control scheme considering the interconnected network switches under self-similar traffic environments.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.11c
• /
• pp.1769-1772
• /
• 2002

iBASE/Cluster는 한국전자통신연구원에서 개발한 객체 지향 멀티미디어 DBMS인 바다-IV를 기반으로 공유 디스크 기반의 클러스터 환경에서 무정지 DBMS 서비스를 지원하기 위하여 확장된 DBMS 이다. 본 논문에서는 이를 위하여 수행된 바다-IV 의 확장에 대해 기술한다. 기술된 내용은 공유 디스크 자원 접근을 동기화 하기 위한 동시성 제어 기법 및 버퍼 관리 기법, 그리고 특정 노드에 고장 발생 시 회복 기법 등에 대한 확장을 포함한다. 또한 확장된 DBMS가 LVS(Linux Virtual Server) 기반의 클러스터에서 자동적인 고장의 감지 및 이전(fail-over)을 수행하기 위한 연동 구조를 다루고 동작 시나리오를 통하여 구현된 시스템의 흐름을 보인다.

• Journal of Korea Multimedia Society
• /
• v.6 no.7
• /
• pp.1219-1230
• /
• 2003

We analyze an integrated voice/data CDMA system, where the whole channels are divided into voice prioritized channels and voice non-prioritized channels. For real-time voice service, a preemptivc priority is granted in the voice prioritized channels. And, for delay-tolerant data service, the employment of buffer is considered. On the other hand, the transmission permission probability in best-effort packet-data service is controlled by estimating the residual capacity available for users. We build a 2-dimensional markov chain about prioritized-voice and stream-data services and accomplish numerical analysis in combination with packet-data traffic based on residual capacity equation.

• Journal of KIISE:Computer Systems and Theory
• /
• v.30 no.10
• /
• pp.523-532
• /
• 2003

Video on Demand(VOD) services cause high resource consumption in a video seuer, because multimedia are characterized by continuous playback, a high bandwidth requirement, and long playback duration. Patching has been proposed to save the network I/O bandwidth of a video server. To achieve true VOD, patching uses multicasting to share video streams, thereby providing immediate VOD services to users without any service latency. A communication channel is used to either multicast the entire video as a regular channel or multicast only the leading portion of a video as a Patching channel. This paper Proposes a buffering patching technique that divides regular channels, as used in patching, into sub-regular channels and regular channels to shorten the holding time of the channels. In the proposed technique, the last portion of video data, corresponding to the size of the buffering window, is not transferred by sub-regular channels, but rather downloaded and buffered in the user buffer from the latest regular channel. When simulations were performed to compare the performance of the proposed technique with that of conventional patching, the results indicated that the proposed technique was superior in terms of the defection rate, average service latency, and fairness, although the amount of video data buffered at each user station was higher than that with patching.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.2 no.1
• /
• pp.87-99
• /
• 1998

In the multicasting communications, the quality of service is not guaranteed occasionally, according to the shortage of resources, because all resources such as communication buffer, processing power for the packet multiplication, and bandwidth, should be shared among each members. This problem can be solved through the hierarchical multicaster construction mechanism which guarantees that all multicaster can be operated within the required performance. This paper proposes the architecture of multipoint communication server based on hypercube type massively parallel computer. Through the simulation, it is verified that the traffic of links between each node of hypercube computer is under the some bandwidth. So, it is certified that the hypercube computer is suitable for the multicast communication servers supporting the various groups.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.18 no.8
• /
• pp.1190-1200
• /
• 1993

In this work, the performance of both delay sensitive traffics and loss sensitive traffics of the output buffered ATM switch with priority control mechanisms has been evaluated. We choose the partial buffer sharing mechanism as the loss priority control mechanism and the HOL(Head Of Line) priority control mechanism as the time priority control mechanism. We model loss sensitive traffics with Poisson process and delay sensitive traffics with MMPP. With loss priority control, it is confirmed that loss probability of loss sensitive traffice decreases when the loss priority control mechanism is chosen. With time priority control, it has also been confirmed that mean cell delay of delay sensitive traffics decreases when the HOL priority control mechanism is used. From this analysis, It has been confirmed that the requirements of QOS for both loss sensitive and delay sensitive traffics can be satisfied in the ATM switch by combining both the loss priority control mechanism and the HOL priority control mechanism.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.9A
• /
• pp.763-773
• /
• 2003

In this paper, we present a register renaming method for conditional execution out-of-order instruction issue microprocessors. Register renaming method reduces false data dependencies (write after read(WAR) and write after write(WAW)). To implement a conditional execution out-of-order instruction issue microprocessor using register renaming, we use a register file which includes both in-order state physical registers and look-ahead state physical registers to share all logical registers. And we design an in-order state indicator, a renaming state indicator, a physical register assigning indicator, a condition prediction buffer and a reorder buffer. As we utilize the above hardwares, we can do register renaming and trace the in-order state. In this paper, we present an improved register renaming method using smaller hardware resources than conventional register renaming method. And this method eliminates an associative lookup and provides a short recovery time.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.33 no.3B
• /
• pp.144-151
• /
• 2008

In this paper, we propose a video streaming server that guarantees a certain level of quality when a server should serve video streaming service to multiple heterogenous clients simultaneously with TCP transport. If each heterogeneous client requests video streaming service in according to its own requirement such as bitrate of content and these requests are accepted by a server, then TCP flows for each video streaming session fairly share limited uplink bandwidth of the server. At this time, because TCP's bandwidth fair-share characteristics can result in bandwidth shrinkage of higher bitrate video streaming session, the client of higher bitrate video may suffer sluggish playback which is related with streaming QoS degradation. To tackle this problem, our proposed server system uses multiple TCP connections adaptively for each video streaming session depending on the anticipated status of the client playout buffer. Simulation results show that our proposed algorithm can successfully reduce the occurrence of playout buffer underrun and enhance streaming quality for whole video clients.