• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.1
• /
• pp.283-290
• /
• 2016

As the development of recent science and technology, high-performance computing resources is needed to solve complex problems. To reach these requirements, it has been actively studied about grid computing that consist of a huge system which bind a heterogeneous high performance computing resources into on which are geographically dispersed. However, The current research situation which are the process to obtain the best results in the limited resources and the scheduling policy to accurately predict the total execution time of the real-time task are very poor. In this paper, in order to overcome these problems, we suggested a grid computing-based risk management system which derived from the system structure and the process for improving the efficiency of the system, grid computing-based working methodology, risk policy module which can manage efficiently the problem of the work of resources(Agent), scheduling technique and allocation method which can re-allocate the resource allocation and the resources in problem, and monitoring which can manage resources(Agent).

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.37 no.2
• /
• pp.44-53
• /
• 2000

The waiting delays of users are inevitable in this policy since the services are not taken immediately upon requests but upon every scheduling points. An inefficient management of such delays makes an unfair service to users and increases the possibility of higher reneging rates. This paper proposes an adaptive batch scheduling scheme which improves the average waiting time of users requests and reduces the starvation problem of users requesting less popular movies. The proposed scheme selects dynamically multiple videos in given intervals based on the service patterns which reflect the popularity distribution(Zipf-distribution) and resource utilizations. Experimental results of simulations show that the proposed scheme improves about 20-30 percentage of average waiting time and reduces significantly the starving requesters comparing with those of conventional methods such as FCFS and MQL.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.4
• /
• pp.87-92
• /
• 2020

Although GPGPU (General-Purpose GPU) can maximize performance by parallelizing a task with tens of thousands of threads, those threads are internally grouped into a thread block, which is a base unit for processing and resource allocation. A thread block scheduler is a specialized hardware gadget whose role is to allocate thread blocks to GPGPU processing hardware in a round-robin manner. However, round-robin is a sequential allocation policy and is not optimized for GPGPU resource utilization. In this paper, we propose a thread block scheduler model which can analyze and quantify performances for various thread block scheduling policies. Experiment results from the implemented simulator of our model show that the legacy hardware thread block scheduling does not behave well when workload becomes heavy.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.2
• /
• pp.385-395
• /
• 1999

Video-On-demand system provides electronic video rental services from remote video servers on a broadband communication networks. When we assume some situation that serve the same video stream by batching, each time single video stream became available at server-end side, we face the problem that what video stream to be scheduled at what time instant. Hence, in order to manage the resource bandwidth which is different at each stream, the policy of channel allocation became important factor. In this work, we proposed three type of scheduling policies with queueing, as are cumulative Waiting Time(CWT), Mean Waiting Time(MWT) and Predictable Popular Factor(PPF). While CWT policy is good for popular video, MWT policy was good for non-popular video. Also, we know that PPT policy has some fairness for average wait time and reneging rate of user requests.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10c
• /
• pp.44-46
• /
• 2000

본 논문은 Tertiary 저장장치를 이용한 대용량 비디오 서버의 설계와 구현에 대해서 기술한다. 서버는 서비스와 프로세스 관리 서브시스템, 자원 관리 서브시스템, 입출력 관리 서브시스템의 세 부분으로 구성된다. 입출력 대역폭이 낮은 Tertiary 저장장치의 효율적인 사용을 위해 디스크 캐쉬 풀 모델을 사용하고, 적절한 스케줄링 정책과 캐슁 알고리즘을 이용하여 가능한 많은 사용자를 수용한다. 서버는 Windows 2000에서 시스템 서비스로 구현되었으며 디스크 캐쉬의 효과적인 사용으로 Tertiary 저장장치의 낮은 대역폭을 최대한 활용하여 접속된 사용자에게 안정된 비디오 스트리밍 서비스를 제공한다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 1998.10a
• /
• pp.3-8
• /
• 1998

주문형 비디오 시스템의 성능에서 중요한 문제는 클라이언트 요구를 만족시키기 위한 입출력 대역폭이다. 이러한 입출력 대역폭을 감소시키기 위해 하나의 비디오 스트림을 사용하여 다중 요청을 처리하는 많은 정책들; 배칭. 피기백 머징, 버퍼링이 연구되고 있다. 본 논문에서는 인기 있는 비디오들에 대한 서비스 요청들이 매 배칭 간격마다 스케줄될 수 있도록 인기 있는 비디오들을 위한 서버 스트림 용량을 동적으로 예약해 두는 예약 기반 동적 배칭 정책을 제안하고, 그것의 성능을 시뮬레이션을 통하여 단순 배칭 정책과 비교하고 평가한다.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.6
• /
• pp.233-241
• /
• 2008

The overlay multicast that has been recently presented as an alternative for the IP multicast needs scheduling method that can apply the system resource and the network bandwidth for the many users. We propose a efficient overlay multicasting network policy for multimedia services with composite stream service topology. And in order to evaluate the ability of the proposed scheme's performance, We test the performance of multi_casting nodes with RTT values, the bandwidth attrition rate for building the network topology and MOR effects with Overcast[5] The result of simulation shows improves about $8{\sim}20%$ of performance.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.10 no.7
• /
• pp.1614-1620
• /
• 2009

The overlay multicast that has been recently presented as an alternative for the IP multicast has been getting much persuasion by the system resource and the network bandwidth and the advancement of the network cost. we propose a overlay multi_casting network policy for efficient multimedia services with P2P join scheduling. In order to evaluate the performance of proposed scheme, We test the performance of multi_casting nodes with average join count, the maximum guaranteed waiting time and buffer starvation ratio with similar techniques. The result of simulation shows improved about 15${\sim}$20% of performance.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.13 no.3
• /
• pp.72-83
• /
• 2017

GPUs can provide high throughput with latency hiding by executing many warps in parallel. MSHR(Miss Status Holding Registers) for L1 data cache tracks cache miss requests until required data is serviced from lower level memory. In recent GPUs, excessive requests for cache resources cause underutilization problem of GPU resources due to cache resource reservation fails. In this paper, we propose a new warp scheduling technique to reduce stall cycles under MSHR resource shortage. Cache miss rates for each warp is predicted based on the observation that each warp shows similar cache miss rates for long period. The warps showing low miss rates or computation-intensive warps are given high priority to be issued when MSHR is full status. Our proposal improves GPU performance by utilizing cache resource more efficiently based on cache miss rate prediction and monitoring the MSHR entries. According to our experimental results, reservation fail cycles can be reduced by 25.7% and IPC is increased by 6.2% with the proposed scheduling technique compared to loose round robin scheduler.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.06b
• /
• pp.493-496
• /
• 2007

그리드 컴퓨팅은 방대한 데이터 저장 공간과 고성능 연산능력을 요구하는 작업에 적합하다. 동적계획법(dynamic programming)은 방대한 크기의 동적 테이블(dynamic table)을 구성하여 최적해(optimal solution)을 찾기 때문에 그리드에서 수행하기에 적합한 작업이다. 본 논문에서는 동적 테이블을 구성하고 분산 배치하는 할당 정책을 제안한다. 그리고 동적계획법 기반의 어플리케이션을 그리드에서 효율적으로 수행할 수 있는 그리드 시스템 구조를 제안한다.