Approximation algorithms for NP-hard problems
Optimal time-critical scheduling via resource augmentation (extended abstract)
STOC '97 Proceedings of the twenty-ninth annual ACM symposium on Theory of computing
STOC '99 Proceedings of the thirty-first annual ACM symposium on Theory of computing
SODA '93 Proceedings of the fourth annual ACM-SIAM Symposium on Discrete algorithms
Speed is as powerful as clairvoyance [scheduling problems]
FOCS '95 Proceedings of the 36th Annual Symposium on Foundations of Computer Science
A unified analysis of hot video schedulers
STOC '02 Proceedings of the thiry-fourth annual ACM symposium on Theory of computing
SODA '03 Proceedings of the fourteenth annual ACM-SIAM symposium on Discrete algorithms
Resource Augmentation in Load Balancing
SWAT '00 Proceedings of the 7th Scandinavian Workshop on Algorithm Theory
Minimizing the Maximum Starting Time On-line
ESA '02 Proceedings of the 10th Annual European Symposium on Algorithms
Minimizing flow time nonclairvoyantly
Journal of the ACM (JACM)
SRPT optimally utilizes faster machines to minimize flow time
SODA '04 Proceedings of the fifteenth annual ACM-SIAM symposium on Discrete algorithms
Theoretical Computer Science - Special issue: Online algorithms in memoriam, Steve Seiden
SRPT optimally utilizes faster machines to minimize flow time
ACM Transactions on Algorithms (TALG)
Scalably scheduling processes with arbitrary speedup curves
SODA '09 Proceedings of the twentieth Annual ACM-SIAM Symposium on Discrete Algorithms
Minimizing the maximum starting time on-line
Information and Computation
Optimal online algorithms on two hierarchical machines with resource augmentation
COCOON'11 Proceedings of the 17th annual international conference on Computing and combinatorics
Scalably scheduling processes with arbitrary speedup curves
ACM Transactions on Algorithms (TALG)
Hi-index | 0.00 |
We consider the problem of preemptive non-clairvoyant scheduling on a single machine. In this model a scheduler receives a number of jobs at different times without prior knowledge of the future jobs or the required processing time of jobs that are not yet completed. We want to minimize the total response time, i.e. the sum of times each job takes from its release to completion.One particular algorithm, Balance, always schedules the job that is least processed so far. A comparison of an on-line scheduler running Balance against the optimal off-line shows a very large competitive ratio if both algorithms use machines of the same speed. However, it has been shown that if Balance is run on a υ times faster machine than its clairvoyant competitor, then the competitive ratio drops to υ/(υ - 1) at most. This result showed that speed can be almost as good as clairvoyance.We show for υ ≥ 2 the competitive ratio of Balance is 2/υ. In other words, sufficiently high speed is more powerful than clairvoyance.