A Survey of Scheduling Results for Imprecise Computation Tasks
Meeting deadline constraints is one of the most important concerns in real-time systems. Sometimes, it is impossible to schedule all of the tasks so that their deadlines are met, a situation that occurs quite often when the system is in peak load. To cope with this situation, one can completely give up certain less important tasks in favor of meeting the deadlines of more important ones. Another approach is to regard each task as logically composed of two subtasks, mandatory and optional; the optional subtask of each task begins after the end of its mandatory subtask. It is required that each task finishes its mandatory subtask completely, while its optional subtask can be left unfinished. If a task cannot finish its optional subtask, it incurs an error proportional to the execution time of its unfinished portion. This approach is particularly useful for iterative algorithms, where the mandatory subtask corresponds to the task of obtaining an initial result and the optional subtask corresponds to the enhancement of previously obtained results. With this approach it is possible to satisfy the deadline constraints of more tasks, even though some tasks may not have completely finished execution.
KeywordsExecution Time Approximation Algorithm Single Processor Total Execution Time Task System
Unable to display preview. Download preview PDF.
- K. I-J. Ho, J. Y-T. Leung, and W-D. Wei, “Scheduling Imprecise Computation Tasks with 0/1 Constraint,” Tech. Rept. UNL-CSE-92-16, Department of Computer Science and Engineering, University of Nebraska-Lincoln, Lincoln, NE, 68588-0115, 1992.Google Scholar
- K. I-J. Ho, J. Y-T. Leung, and W-D. Wei, “Minimizing Maximum Weighted Error for Imprecise Computation Tasks,” J. of Algorithms, to appear.Google Scholar
- K. I-J. Ho, J. Y-T. Leung, and W-D. Wei, “Minimizing Constrained Maximum Weighted Error for Doubly Weighted Tasks,” Tech. Rept. UNL-CSE-92-18, Department of Computer Science and Engineering, University of Nebraska-Lincoln, Lincoln, NE, 68588-0115, 1992.Google Scholar
- J. Y-T. Leung, T. W. Tarn, and C. S. Wong, “Heuristics for Minimizing Mean Flow Time with Error Constraint,” Tech. Rept. UTDCS-25-89, Computer Science Program, University of Texas at Dallas, Richardson, TX, 75083, 1989.Google Scholar
- J. Y-T. Leung, T. W. Tarn, C. S. Wong, and G. H. Young, “Minimizing Mean Flow Time with Error Constraint,” Proc. of the 10th Real-Time Systems Symposium, pp. 2–11, Santa Monica, CA, 1989.Google Scholar
- J. Y-T. Leung, V. K. M. Yu, and W-D. Wei, “Minimizing the Weighted Number of Tardy Task Units,” Discrete Applied Math., to appear.Google Scholar
- K-J. Lin, S. Natarajan, and J. W-S. Liu, “Concord: A Distributed System Making Use of Imprecise Results,” Proc. of COMPSAC’ 87, Tokyo, Japan, Oct. 1987.Google Scholar
- K-J. Lin, S. Natarajan, and J. W-S. Liu, “Imprecise Results: Utilizing Partial Computations in Real-Time Systems,” Proc. of the 8th Real-Time Systems Symposium, San Francisco, CA, Dec. 1987.Google Scholar
- K-J. Lin, S. Natarajan, and J. W-S. Liu, “Scheduling Real-Time, Periodic Job Using Imprecise Results,” Proc. of the 8th Real-Time Systems Symposium, San Francisco, CA, Dec. 1987.Google Scholar
- J. W-S. Liu, K-J. Lin, W-K. Shih, A. C. Yu, J-Y. Chung, and W. Zhao, “Algorithms for Scheduling Imprecise Computations,” IEEE Computer, 24 (1991), pp. 58–68.Google Scholar
- J. B. Orlin, “A Faster Strongly Polynomial Minimum Cost Flow Algorithm,” Proc. of the 20th ACM Symposium on Theory of Computing, 1988, pp. 377–387.Google Scholar
- W-K. Shih, J. W-S. Liu, J-Y. Chung, and D. W. Gillies, “Scheduling Tasks with Ready Times and Deadlines to Minimize Average Error,” ACM Operating Systems Review, July 1989.Google Scholar