Large-scale distributed computing systems divide a job into many independent tasks and run them in parallel on differ- ent machines. A challenge in such parallel computing is that the time taken by a machine to execut...
详细信息
ISBN:
(纸本)9781450327893
Large-scale distributed computing systems divide a job into many independent tasks and run them in parallel on differ- ent machines. A challenge in such parallel computing is that the time taken by a machine to execute a task is inherently variable, and thus the slowest machine becomes the bottle- neck in the completion of the job. One way to combat the variability in machine response is to replicate tasks on multi- ple machines and waiting for the machine that finishes first. While task replication reduces response time, it generally increases resource usage. In this work, we propose a theo- retical framework to analyze the trade-off between response time and resource usage. Given an execution time distribu- tion for machines, our analysis gives insights on when and how replication helps. We also propose efficient scheduling algorithms for large-scale distributed computing systems.
Coarse-Grained Reconfigurable Arrays (CGRAs) exhibit high performance, improved flexibility, low cost, and power efficiency for various application domains. Compute-intensive loop kernels, which are perfect candidates...
详细信息
Round-robin based packet scheduling algorithms are suitable for high-speed networks, but in a variable-length packet environment, this kind of scheduling algorithm need consider packet length to guarantee scheduling f...
详细信息
The traffic of multimedia service in wireless network increases by exponential as the number of applications increase. Since the weakness of proportional fair algorithm in meeting the Quality of Service (QoS) such as ...
详细信息
In this paper, we study non-preemptive uniprocessor realtime scheduling using the non-preemptive RM (npRM) and EDF (npEDF) scheduling algorithms. We discuss the limitations of existing studies, identifying pessimism i...
详细信息
CoMP (Coordinated Multiple Point transmission/reception) is recognized as a promising technology to increase the cell edge user throughput and spectral efficiency in LTE-Advanced. This paper proposes a muti-user sched...
详细信息
CoMP (Coordinated Multiple Point transmission/reception) is recognized as a promising technology to increase the cell edge user throughput and spectral efficiency in LTE-Advanced. This paper proposes a muti-user scheduling algorithm considering fairness and cell order selection, where primary user is selected by considering objective fairness instead of proportional fairness and cell order selection is selected by comparing user set capacity rather randomly. Experiment results in MATLAB platform demonstrate that the proposed algorithm can be better to improve fairness and spectral efficiency than algorithm based on channel correlation.
We describe a multi-objective evolutionary algorithm that derives schedules for the English Football League over the busy New Year period according to seven objectives. The two principal objectives are to minimise tra...
详细信息
A new fair subcarrier scheduling algorithm is proposed for the OFDMA relay system to achieve the balance between the system throughput and the user fairness, rather than merely to pursue the maximum of the system thro...
详细信息
Synchronization of task usually happens in real-time embedded systems. But it is not easy to achieve a good performance for the synchronization tasks, especially on multicore processor. After analysis of the task char...
详细信息
Synchronization of task usually happens in real-time embedded systems. But it is not easy to achieve a good performance for the synchronization tasks, especially on multicore processor. After analysis of the task characteristics on homogeneous multicore processor, we proposed a Longest Synchronization-Point Time (LSPT) task scheduling algorithm to obtain the good real-time performance and the minimum completing execution time of synchronization tasks. In the LSPT, we selected the most suitable task to run on the kernel based-on priority and execution time of synchronization point of tasks, to make the group of synchronization tasks to get the minimum completion time. In the last, we test the realtime performance of LSPT, the experimental results demonstrate the LSPT algorithm can achieve good efficiency for parallel synchronization tasks on multicore processor.
First of all, in view of the process industry production scheduling system, the mathematical model of the process industry production scheduling was established. Secondly, the shortest maximum completion time was take...
详细信息
暂无评论