• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04a
• /
• pp.76-78
• /
• 2002

최근 멀티미디어 데이터 서버로 리눅스 시스템을 쓰는 경우가 많아 졌다. 이 경우, 멀티미디어 데이터에 관한 서비스를 효율적으로 제공할 수 있어야 한다. 이를 위하여, 본 논문에서는 기존의 리눅스 scheduling 방식이 갖는 단일레벨 run-queue 구조를 변형한 다중레벨 run-queue를 제안하였다. 기존의 단일레벨 run-queue에서는 queue 내에 프로세스의 수가 많아질수록 검색시간이 길어지는 단점이 있다. 본 논문에서는 기존의 run-queue를 여러 단계로 나누고, scheduling 과정에서 상위 queue부터 프로세스가 존재하는지를 조사하도록 스케줄러를 변형하였다. 따라서, 상위 queue에 프로세스가 있는 경우에는 하위의 queue는 더 이상 조사한 필요가 없게 되므로. 결과적으로 검색시간을 줄일 수 있게 된다. 한편, 다중 레벨의 run-queue를 사용할 경우, 이를 관리하기 위한 오버헤드가 별도로 발생한다. 본 논문에서는, 제안한 다중레벨의 run-queue 시스템의 성능을 최적화하기 위하여, queue의 적절한 개수 선정 및 각 프로세스를 어떤 queue에 넣을 것인지를 결정하는 것이 성능에 미치는 영향에 대하여 실험적으로 연구하여 새로운 스케줄러의 성능을 기존 스케줄러와 비로 분석하였다. 본 논문에서 제안한 멀티레벨 run-queue를 사용함으로써, 각 queue의 스케줄링 정책(policy)과 관련 파라메터 간을 독자적으로 변화시킬 수 있다. 따라서, 여러 가지 상황에 적합한 스케줄링을 각각의 경우에 맞게 최적화하는 것이 손쉬워 지므로 여러 분야에서 매우 유용하게 쓰일 것이다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10c
• /
• pp.100-102
• /
• 2000

본 논문에서는 디스크의 이용률을 향상시키기 위하여 효율적인 디스크 스케줄링 알고리즘을 제안한다. 디스크 요구는 제공되는 서비스의 레벨을 따라 주기적인 요구, 비 주기적인 요구, 대화형 요구로 나눈다. 기존의 멀티미디어 서비스를 위한 디스크 스케줄링은 공정성 및 디스크 대역폭 보장만을 고려하였는데 본 논문에서는 대화형 요구를 고려하여 디스크 스케줄링의 공정성 및 주기적인 요구의 대역폭을 보장하고 디스크 이용률을 향상시켜 처리 가능한 요구 수를 증가시킬 수 있는 디스크 스케줄링을 제안한다.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.47 no.12
• /
• pp.61-68
• /
• 2010

In this paper, we investigate the performance of ARQ-aided Time Switched Transmit Diversity (ARQ-TSTD) applying the multi-level control signaling in the WCDMA LCR-TDD system. Proposed ARQ-TSTD system applies the multi-level control signaling scheme in which the receiver sends the response signal (ACK or NACK signal) to the transmitter and defines NACK2 signal for multi-level control. Transmitter utilize the NACK2 control signal to the postponement of transmission and multi-user scheduling scheme proposed by this paper. Simulation results demonstrate that the proposed postponement of transmission and multi-user scheduling scheme yield about 1.3dB, 1.4dB performance gain respectively, compared with the conventional ARQ-TSTD with antenna switching scheme in tenn of the frame error rate (FER) for mobile speed of 3km/h and FER value of 10%. In addition, 14% and 11.5% of throughput gain respectively is shown when Eb/N0=-3dB.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.3A
• /
• pp.318-323
• /
• 2004

In this paper, we present a reducing power consumption of a scheduling for reuse module selection under the time constraint. Traditional high-level synthesis do not allow reuse of complex, realistic datapath component during the task of scheduling. On the other hand, the proposed scheduling of reducing power consumption is able to approach a productivity of the design the low power to reuse which given a library of user-defined datapath component and to share of resource sharing on the switching activity in a shared resource. Also, we are obtainable the optimal the scheduling result in experimental results of our approach various HLS benchmark environment using chaining and multi-cycling in the scheduling techniques.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.3
• /
• pp.161-166
• /
• 2006

In this paper, CPLD low power technology mapping using reuse module selection under the time constraint is proposed. Traditional high-level synthesis do not allow reuse of complex, realistic datapath component during the task of scheduling. On the other hand, the proposed algorithm is able to approach a productivity of the design the low power to reuse which given a library of user-defined datapath component and to share of resource sharing on the switching activity in a shared resource Also, we are obtainable the optimal the scheduling result in experimental results of our using chaining and multi-cycling in the scheduling techniques. Low power circuit make using CPLD technology mapping algorithm for selection reuse module by scheduling.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.11A
• /
• pp.1106-1112
• /
• 2007

As a demand for high speed uplink packet services increases, the WCDMA enhanced uplink, also known as high speed uplink packet access (HSUPA), has been specified in release 6 by 3GPP. This HSUPA will provide various types of multimedia services, such as real-time video streaming, gaming, VoIP, and FTP. Generally, the performance of HSUPA is dominated by scheduling policy. Therefore, it is required to design a scheduling algorithm considering the traffic characteristics to provide QoS guaranteed services in various traffic environments. In this paper, we propose a scheduling algorithm considering the traffic characteristics to guarantee QoS in a mixed traffic environment. Finally, the performance of the proposed scheduling algorithm is evaluated in terms of average packet delay, packet delay jitter, and system throughput using a system level simulator.

• KIPS Transactions on Computer and Communication Systems
• /
• v.6 no.5
• /
• pp.219-230
• /
• 2017

General-Purpose Graphics Processing Units (GPGPUs) build massively parallel architecture and apply multithreading technology to explore parallelism. By using programming models like CUDA, and OpenCL, GPGPUs are becoming the best in exploiting plentiful thread-level parallelism caused by parallel applications. Unfortunately, modern GPGPU cannot efficiently utilize its available hardware resources for numerous general-purpose applications. One of the primary reasons is the inefficiency of existing warp/thread block schedulers in hiding long latency instructions, resulting in lost opportunity to improve the performance. This paper studies the effects of hardware thread scheduling policy on GPGPU performance. We propose a novel warp scheduling policy that can alleviate the drawbacks of the traditional round-robin policy. The proposed warp scheduler first classifies the warps of a thread block into two groups, warps with long latency and warps with short latency and then schedules the warps with long latency before the warps with short latency. Furthermore, to support the proposed warp scheduler, we also propose a supplemental technique that can dynamically reduce the number of streaming multiprocessors to which will be assigned thread blocks when encountering a high contention degree at the memory and interconnection network. Based on our experiments on a 15-streaming multiprocessor GPGPU platform, the proposed warp scheduling policy provides an average IPC improvement of 7.5% over the baseline round-robin warp scheduling policy. This paper also shows that the GPGPU performance can be improved by approximately 8.9% on average when the two proposed techniques are combined.

• Journal of IKEEE
• /
• v.7 no.2 s.13
• /
• pp.166-171
• /
• 2003

In this paper, we propose a minimal power data path allocation algorithm for low power circuit design. The proposed algorithm minimizes switching activity for input variables in scheduled CDFG. Allocations are further divided into the tasks of register allocation and module allocation. The register allocation algorithm execute that it eliminate spurious switching activity in functional unit and minimize the numbers of multiplexer. Also, resource allocation method selects a sequence of operations for a module such that the switching activity is reduced. Therefore, the algorithm executes to minimize the switching activity of input values, sequence of operations and number of multiplexer. Experimental results using benchmarks show that power is reduction effect from 13% to 17% power consumption, when compared with the Genesis-lp high-level synthesis system.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.05a
• /
• pp.123-124
• /
• 2013

비용대비 작업 처리의 효율을 위해서는 사용자들의 작업 요구사항에 적절한 자원을 선택하고 요구 작업을 적절한 할당된 자원에 스케쥴링하는 플랫폼이 필수적이다. 또한 이러한 플랫폼은 사용자의 SLA 에 따라 작업 처리 기한 안에 요구 비용 이내로 작업을 처리할 수 있도록 결정을 내릴 수 있어야 하고 요구 작업량의 변화에 따라 즉각 대응을 하기 위하여 실시간적인 결정을 내릴 수 있어야 한다. 이러한 복잡한 결정 사항들을 최적 판단으로 대신 처리해주는 미들웨어로 클라우드 자원 브로커 시스템을 사용할 수 있다. 클라우드 자원 브로커 시스템은 작업 스케쥴링과 자원 프로비저닝 등이 가격, 처리시간에 중요한 선택 및 수행을 한다. 기존의 많은 논문들에서의 작업 스케줄링은 다중 테넌트 정책의 클라우드가 제공하는 사용자들간의 가상 머신 독립에 초점을 두어 하나의 가상 머신이 하나의 작업에 한정되도록 처리하는 방식이었다. 이는 병렬화의 정도가 낮은 어플리케이션의 경우 시스템 활용률이 낮아 자원 활용율이 떨어진다. 이를 다수의 작업을 멀티 태스킹, 멀티 스레드의 방법으로 하나의 가상 머신에서 처리하도록 하여 스레드 레벨 병렬화의 이점을 이용해 자원 이용률을 높임으로 효율을 높이고자 한다.

• Journal of KIISE
• /
• v.43 no.6
• /
• pp.613-620
• /
• 2016

This paper presents a MOHA(Many-Task Computing on Hadoop) framework which aims to effectively apply the Many-Task Computing(MTC) technologies originally developed for high-performance processing of many tasks, to the existing Big Data processing platform Hadoop. We present basic concepts, motivation, preliminary results of PoC based on distributed message queue, and future research directions of MOHA. MTC applications may have relatively low I/O requirements per task. However, a very large number of tasks should be efficiently processed with potentially heavy inter-communications based on files. Therefore, MTC applications can show another pattern of data-intensive workloads compared to existing Hadoop applications, typically based on relatively large data block sizes. Through an effective convergence of MTC and Big Data technologies, we can introduce a new MOHA framework which can support the large-scale scientific applications along with the Hadoop ecosystem, which is evolving into a multi-application platform.