• 한국통신학회논문지
• /
• 제38C권6호
• /
• pp.521-530
• /
• 2013

본 논문은 기존의 소프트웨어가 멀티코어기반의 모바일 디바이스를 인지할 수 있도록 태스크 병렬화 시스템(BioMP)을 제안한다. 애플리케이션 개발자가 기존의 소프트웨어에 병렬화 규약의 코드를 추가하였을 때, 제안 시스템은 호환성 뿐만 아니라 병렬 쓰레드의 수행을 지원한다. BioMP는 기존의 대용량 애플리케이션 소스코드를 단시일에 멀티코어를 인지하는 소프트웨어로 개선한다. 실험 결과, 우리의 아이디어는 쿼드 코어기반의 멀티코어 환경에서 기존의 시스템 대비 애플리케이션 실행속도를 약 64%까지 개선하였다. 또한, BioMP는 독립적인 컴포넌트이기 때문에 어떠한 플랫폼의 추가적인 수정도 필요로 하지 않는다. 그 결과, 애플리케이션 개발자는 멀티코어향 소프트웨어를 애플리케이션 스토어에 배포하였을 때, 사용자는 모바일 디바이스의 어떠한 수정도 없이 즉시 실행을 할 수 있다.

• 한국안전학회지
• /
• 제13권4호
• /
• pp.230-239
• /
• 1998

Though limited, Model Human Processor (MHP) has been used to explain the complex users' behaviors during human-computer interactions in a simplified manner. MHP consists of perceptual, cognitive and motor systems, each with processors and memories interacting with each other in serial or parallel mode. The important parameters of memory include the storage capacity, the decay time, and the code type of a memorized item. The important parameter of a processor is the cycle time. Using these features of the model, this study developed a computerized cognitive process simulator to predict the cognitive process time of a class match task process. An experimental validity test result showed that the mean prediction time for cognitive process of the class match task simulated 50 times by the simulator was consistent with the mean cognitive process time of the same task performed by 37 subjects. Animation of the data flow during the class match task simulation will help understand the invisible human cognitive process.

• ETRI Journal
• /
• 제45권1호
• /
• pp.93-104
• /
• 2023

Coreference resolution is a task in discourse analysis that links several headwords used in any document object. We suggest pointer networks-based coreference resolution for Korean using multi-task learning (MTL) with an attention mechanism for a hierarchical structure. As Korean is a head-final language, the head can easily be found. Our model learns the distribution by referring to the same entity position and utilizes a pointer network to conduct coreference resolution depending on the input headword. As the input is a document, the input sequence is very long. Thus, the core idea is to learn the word- and sentence-level distributions in parallel with MTL, while using a shared representation to address the long sequence problem. The suggested technique is used to generate word representations for Korean based on contextual information using pre-trained language models for Korean. In the same experimental conditions, our model performed roughly 1.8% better on CoNLL F1 than previous research without hierarchical structure.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1988년도 한국자동제어학술회의논문집(국내학술편); 한국전력공사연수원, 서울; 21-22 Oct. 1988
• /
• pp.1-5
• /
• 1988

The realization of high quality robot control needs the improvement of computing speed of controller. In this paper, parallel processing method is considered for this purpose. A S/W algorithm for task scheduling is developed first, and then, an appropriate H/W structure is proposed. This scheme is applied to calculate inverse kinematics of PUMA robot. The simulation results show that the computing time when using three 8086/87's is reduced to 4.23 msec compared to 10 msec in case using one 8086/87.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2003년도 가을 학술발표논문집 Vol.30 No.2 (3)
• /
• pp.631-633
• /
• 2003

인터넷 가반 분산/병렬 처리 프레임 워크 PDP(Parallel/Distributed Processing Scheme on Web)는 네트워크 내 유휴 상태 호스트들을 활용하여 대용량 작업을 병렬로 처리한다. 본 논문에서는 이러한 서브 작업을 할당받는 자원이 동작하는 네트워크 환경을 모니터링 함으로써 수시로 변화하는 네트워크 환경에 대처하는 방안을 제시한다. 특히 네트워크 환경 모니터링 예측 결과를 PDP의 작업 할당 알고리즘에 적용하여 네트워크 과부하 및 결함 등으로 인해 발생되는 작업 지연 요소에 적응적 대처함으로써 전체 작업 수행 처리율 향상을 도모하는 방법을 제안한다.

• ETRI Journal
• /
• 제42권5호
• /
• pp.748-760
• /
• 2020

MapReduce is a framework that can process huge datasets in parallel and distributed computing environments. However, a single machine failure during the runtime of MapReduce tasks can increase completion time by 50%. MapReduce handles task failures by restarting the failed task and re-computing all input data from scratch, regardless of how much data had already been processed. To solve this issue, we need the computed key-value pairs to persist in a storage system to avoid re-computing them during the restarting process. In this paper, the task failure resilience (TFR) technique is proposed, which allows the execution of a failed task to continue from the point it was interrupted without having to redo all the work. Amazon ElastiCache for Redis is used as a non-volatile cache for the key-value pairs. We measured the performance of TFR by running different Hadoop benchmarking suites. TFR was implemented using the Hadoop software framework, and the experimental results showed significant performance improvements when compared with the performance of the default Hadoop implementation.

• 한국통신학회논문지
• /
• 제25권5B호
• /
• pp.899-912
• /
• 2000

PVM(Parallel Virtual Machine)은 네트워크 상에 분산되어 있는 여러 시스템들을 투명하게 활용하여 사용자에 고성능 병렬 컴퓨팅을 지원하는 단일 가상 시스템(single virtual system)으로 인식되게 하는 middle-ware 소프트웨어이다. 초고속 통신망을 기반으로 하는 PVM에서는 임의의 태스크로부터 하나의 메시지를 전송하기 위해 세 번의 메시지 복사가 필요하며, 이는 PVM의 성능 저하를 가져오는 주요 원인이 된다. 이러한 문제점을 개선하기 위해 이 논문에서는 zero-copy PVM 통신 모델을 제안한다. 제안된 모델은 PVM 태스크, PVM 데몬, 네트워크 인터페이스 보드 등에 의해 동시 접근이 가능한 전역 공유 메모리(global shared memory)를 이용하며, 초고속 통신망을 기반으로 하고 있다. 이 모델에서 PVM 태스크는 전송하고자 하는 메시지를 전역 공유 메모리에 저장하고, 메시지를 보낼 준비가 되었음을 PVM 데몬에게 알리며, 데몬은 해당 메시지를 커널을 통하지 않고 바로 초고속 통신망으로 전송함으로써, 메시지의 메모리 복사 횟수를 감소 시킨다. 실험 결과 두 시스템간의 메시지 왕복 시간은 제안된 모델을 사용함으로써 현저히 줄어 들었음을 확인하였다.

• 한국시뮬레이션학회:학술대회논문집
• /
• 한국시뮬레이션학회 1994년도 추계학술발표회 및 정기총회
• /
• pp.32-32
• /
• 1994

In this example, a message passing based multicomputer system with general interdonnedtion network is considered. After multicomputer systems are developed with morm-hole routing network, topologies of interconecting network are not major considertion for process management and resource sharing. Tehre is an independeent operating system kernel oneach node. It communicates with other kernels using message passingmechanism. Based on this architecture, the problem is how mech does performance degradation will occur in the case of processor sharing on multicomputer systems. Processor sharing between application programs is veryimprotant decision on system performance. In almost cases, application programs running on massively parallel computer systems are not so much user-interactive. Thus, the main performance index is system throughput. Each application program has various communication patterns. and the sharing of processors causes serious performance degradation in hte worst case such that one processor is shared by two processes and another processes are waiting the messages from those processes. As a result, considering this problem is improtant since it gives the reason whether the system allows processor sharingor not. Input data has many parameters in this simulation . It contains the number of threads per task , communication patterns between threads, data generation and also defects in random inupt data. Many parallel aplication programs has its specific communication patterns, and there are computation and communication phases. Therefore, this phase informatin cannot be obtained random input data. If we get trace data from some real applications. we can simulate the problem more realistic . On the other hand, simualtion results will be waseteful unless sufficient trace data with varisous communication patterns is gathered. In this project , random input data are used for simulation . Only controllable data are the number of threads of each task and mapping strategy. First, each task runs independently. After that , each task shres one and more processors with other tasks. As more processors are shared , there will be performance degradation . Form this degradation rate , we can know the overhead of processor sharing . Process scheduling policy can affects the results of simulation . For process scheduling, priority queue and FIFO queue are implemented to support round-robin scheduling and priority scheduling.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권4호
• /
• pp.1100-1122
• /
• 2023

Scheduling user-submitted cloud tasks to the appropriate virtual machine (VM) in cloud computing is critical for cloud providers. However, as the demand for cloud resources from user tasks continues to grow, current evolutionary algorithms (EAs) cannot satisfy the optimal solution of large-scale cloud task scheduling problems. In this paper, we first construct a large- scale multi-objective cloud task problem considering the time and cost functions. Second, a multi-objective optimization algorithm based on multi-factor optimization (MFO) is proposed to solve the established problem. This algorithm solves by decomposing the large-scale optimization problem into multiple optimization subproblems. This reduces the computational burden of the algorithm. Later, the introduction of the MFO strategy provides the algorithm with a parallel evolutionary paradigm for multiple subpopulations of implicit knowledge transfer. Finally, simulation experiments and comparisons are performed on a large-scale task scheduling test set on the CloudSim platform. Experimental results show that our algorithm can obtain the best scheduling solution while maintaining good results of the objective function compared with other optimization algorithms.

• 제어로봇시스템학회논문지
• /
• 제6권1호
• /
• pp.81-90
• /
• 2000

Parallel manipulator is suitable for the high precise task because it than has higher stiffness, larger load capacity and more excellent precision, due to the closed-lop structure, than serial manipulator. But the controller design for parallel manipulator is difficult because the parallel manipulator has both the complexity of structure and the interference of actuators. The precision improvement of parallel manipulator using a classical linear control scheme is difficult because the parallel manipulator has the tough nonlinear characteristics. In this paper, firstly, the kinematic analysis of a parallel manipulator used at the experiments is performed so as to show the controllability. The analysis of internal singular configuration of the workspace is performed using the kinematic isotropic index so a sto show the limitation of control performance of a simple linear controller with fixed control gains. Secondly, a control scheme is designed by using an adaptive fuzzy logic controller so that active joints of the parallel manipulator track more precisely the desired input trajectory. This adaptive fuzzy logic controller so that active joints of the parallel manipulator track more precisely the desired input trajectory. This adaptive fuzzy logic controller is often used for the control of nonlinear system because it has both the inference ability and the learning ability. Lastly, the effeciency of designed control scheme is demonstrated by the real-time control experiments with IBM PC interface logic H/W and S/W of my won making. The experimental results was a success.