• 한국컴퓨터정보학회논문지
• /
• 제24권4호
• /
• pp.1-9
• /
• 2019

LRR(Loose Round Robin) warp scheduling policy for GPU architecture results in high warp-level parallelism and balanced loads across multiple warps. However, traditional LRR policy makes multiple warps execute long latency operations at the same time. In cases that no more warps to be issued under long latency, the throughput of GPUs may be degraded significantly. In this paper, we propose a new warp scheduling policy which utilizes latency hiding, leading to more utilized memory resources in high performance GPUs. The proposed warp scheduler prioritizes memory instruction based on GTO(Greedy Then Oldest) policy in order to provide reduced memory stalls. When no warps can execute memory instruction any more, the warp scheduler selects a warp for computation instruction by round robin manner. Furthermore, our proposed technique achieves high performance by using additional information about recently committed warps. According to our experimental results, our proposed technique improves GPU performance by 12.7% and 5.6% over LRR and GTO on average, respectively.

• 한국정보과학회논문지:시스템및이론
• /
• 제37권2호
• /
• pp.76-89
• /
• 2010

분산 이기종 시스템의 성능은 DAG로 주어지는 입력 그래프를 스케줄링 하는 알고리즘의 성능에 좌우된다. 많은 스케줄링 알고리즘 중에 리스트 스케줄링 알고리즘은 낮은 복잡도를 가지면서 우수한 성능을 보이고 있다. 리스트 스케줄링은 태스크 우선순위 결정 단계와 프로세서 할당 단계로 이루어져 있으나 대부분의 연구들은 태스크 우선순위 결정 단계만을 연구하고 있다. 본 논문에서는 기존의 할당 정책과 동일한 복잡도를 가지면서 성능이 향상된 새로운 프로세서 할당 정책인 LIP 정책을 제안한다. 기존의 리스트 스케줄링 알고리즘인 HEFT, HCPT, GCA, PETS의 태스크 우선순위 결정 정책에 본 논문에서 제안한 LIP 정책을 적용하여 실험한 결과 기존의 프로세서 할당 정책인 삽입 정책과 비 삽입 정책보다 성능 향상이 있는 것을 확인할 수 있었다.

• 한국시뮬레이션학회논문지
• /
• 제8권3호
• /
• pp.67-76
• /
• 1999

The study of the optimization of operating policy of AGV system, which is used in many factory automation environments has been proceeded by many researchers. The major operating policy of AGV system consists of routing and scheduling policy. AGV routing is composed with collision avoidance and minimal cost path find algorithm. To allocate jobs to the AGV system, AGV scheduling has to include AGV selection rules, parking rules, and recharging rules. Also in these rules, the key time parameters such as processing time of the device, loading/unloading time and charging time should be considered. In this research, we compare and analyze several operating policies of multiple loop-multiple AGV system by making a computer model and simulating it to present an appropriate operating policy.

• 한국경영과학회지
• /
• 제13권1호
• /
• pp.57-57
• /
• 1988

This paper develops an adaptive scheduling policy for flexible manufacturing systems. The inductive learning methodology used for constructing this state-dependent scheduling policy provides and understanding of the relative importance of the various system parameters in determining the appropriate scheduling rule. Experimental studies indicated the superiority of the suggested approach over the alternative approach involving the repeated application of a single scheduling rule for randomly generated test problems as well as a real system, and under both stationary and nonstationary conditions. In particular, its relative performance improves further when there are frequent disruptions, and when disruptions are caused by the introduction of tiiight due date jobs, one of the most common surces of disruptions in most manufacturing systems.

• 경영과학
• /
• 제13권1호
• /
• pp.57-70
• /
• 1996

This paper develops an adaptive scheduling policy for flexible manufacturing systems. The inductive learning methodology used for constructing this state-dependent scheduling policy provides and understanding of the relative importance of the various system parameters in determining the appropriate scheduling rule. Experimental studies indicated the superiority of the suggested approach over the alternative approach involving the repeated application of a single scheduling rule for randomly generated test problems as well as a real system, and under both stationary and nonstationary conditions. In particular, its relative performance improves further when there are frequent disruptions, and when disruptions are caused by the introduction of tiiight due date jobs, one of the most common surces of disruptions in most manufacturing systems.

• Journal of Communications and Networks
• /
• 제7권4호
• /
• pp.408-416
• /
• 2005

Quality of service (QoS) provision is an important and indispensable function for multi-service wireless networks. In this paper, we present a new scheduling/admission control frame­work, including an efficient rate-guaranteed opportunistic scheduling (ROS) scheme and a coordinated admission control (ROS­CAC) policy to support statistic QoS guarantee in multi-service wireless networks. Based on our proposed mathematical model, we derive the probability distribution function (PDF) of queue length under ROS and deduce the packet loss rate (PLR) for individual flows. The new admission control policy makes admission decision for a new incoming flow to ensure that the PLR requirements of all flows (including the new flow) are satisfied. The numerical results based on ns-2 simulations demonstrate the effectiveness of the new joint packet scheduling/admission control framework.

• 반도체디스플레이기술학회지
• /
• 제23권2호
• /
• pp.107-113
• /
• 2024

Scheduling of semiconductor test facilities has been crucial since effective scheduling contributes to the profits of semiconductor enterprises and enhances the quality of semiconductor products. This study aims to solve the scheduling problems for the wafer burn-in test facilities of the semiconductor back-end process by utilizing simulation and deep reinforcement learning-based methods. To solve the scheduling problem considered in this study. we propose novel state, action, and reward designs based on the Markov decision process. Furthermore, a neural network is trained by employing the recent RL-based method, named proximal policy optimization. Experimental results showed that the proposed method outperformed traditional heuristic-based scheduling techniques, achieving a higher due date compliance rate of jobs in terms of total job completion time.

• 산업경영시스템학회지
• /
• 제15권26호
• /
• pp.21-31
• /
• 1992

Today, manpower is used as a major factor in the production of service industry that is on the increasing trend. So the research on manpower structure have been regarded as a important field. The research on personnel scheduling, however, have got de-emphasizeed as compared wi th the field of manufacturing scheduling. As the increasing trend of service industry, an appropriate personnel scheduling for each type of service industry is needed urgently. This paper deals with nurse scheduling problem in the field of personnel scheduling. We aim at developing prototype of the expert nurse scheduling program which is easy to use when scheduling without help of expert, and satisfies the various request of hospital scheduling policy and nurse scheduling.

• 산업공학
• /
• 제7권2호
• /
• pp.107-120
• /
• 1994

This study introduces a scheduling policy to achieve unmanned operations in machining shops representing the feature of both FMS and machining centers. The proposed policy considers the practical constraints in the machining shops such as lot sizes, multiple pallets, and jig/fixtures. In addition, rescheduling to cope with extraordinary events(eg., arrival of high-priority jobs and breakdown of machines) is included. Basically, the scheduling policy consists of four procedures; (1) selection of a job according to dispatching rule, (2) selection of a machine, (3) assignment of jobs to a multiple pallet for the unmanned operation during a certain shift, and (4) prevention of unexpected machine idle caused by constraints on jig/fixture.

• 한국경영과학회지
• /
• 제27권2호
• /
• pp.33-49
• /
• 2002

CBM (Condition-Based Maintenance) has increasingly drawn attention in industry because of its many benefits. CBM Problem Is characterized as a state-dependent scheduling model that demands simultaneous maintenance actions, each for an attribute that influences on machine condition. This problem is very hard to solve within conventional Markov decision process framework. In this paper, we present an intelligent machine maintenance scheduler, for which a new incremental decision tree learning method as evolutionary system identification model and shortest path problem as schedule generation model are developed. Although our approach does not guarantee an optimal scheduling policy in mathematical viewpoint, we verified through simulation based experiment that the intelligent scheduler is capable of providing good scheduling policy that can be used in practice.