TY - GEN
T1 - Resource augmentation in load balancing
AU - Azar, Yossi
AU - Epstein, Leah
AU - Van Stee, Rob
N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2000.
PY - 2000
Y1 - 2000
N2 - We consider load balancing in the following setting. The on-line algorithm is allowed to use n machines, whereas the optimal off-line algorithm is limited to m machines, for some ffixed m < n. We show that while the greedy algorithm has a competitive ratio which decays linearly in the inverse of n=m, the best on-line algorithm has a ratio which decays exponentially in n=m. Specifically, we give an algorithm with competitive ratio of 1 + 1/2n/m(1−o(1)), and a lower bound of 1 + 1/en/m(1+o(1)) on the competitive ratio of any randomized algorithm. We also consider the preemptive case.We show an on-line algorithm with a competitive ratio of 1 + 1=en/m(1+o(1)). We show that the algorithm is optimal by proving a matching lower bound. We also consider the non-preemptive model with temporary tasks. We prove that for n = m + 1, the greedy algorithm is optimal. (It is not optimal for permanent tasks).
AB - We consider load balancing in the following setting. The on-line algorithm is allowed to use n machines, whereas the optimal off-line algorithm is limited to m machines, for some ffixed m < n. We show that while the greedy algorithm has a competitive ratio which decays linearly in the inverse of n=m, the best on-line algorithm has a ratio which decays exponentially in n=m. Specifically, we give an algorithm with competitive ratio of 1 + 1/2n/m(1−o(1)), and a lower bound of 1 + 1/en/m(1+o(1)) on the competitive ratio of any randomized algorithm. We also consider the preemptive case.We show an on-line algorithm with a competitive ratio of 1 + 1=en/m(1+o(1)). We show that the algorithm is optimal by proving a matching lower bound. We also consider the non-preemptive model with temporary tasks. We prove that for n = m + 1, the greedy algorithm is optimal. (It is not optimal for permanent tasks).
UR - http://www.scopus.com/inward/record.url?scp=84949815279&partnerID=8YFLogxK
U2 - 10.1007/3-540-44985-X_17
DO - 10.1007/3-540-44985-X_17
M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???
AN - SCOPUS:84949815279
SN - 3540676902
SN - 9783540676904
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 189
EP - 199
BT - Algorithm Theory - SWAT 2000
A2 - Halldórsson, Magnús M.
PB - Springer Verlag
Y2 - 5 July 2000 through 7 July 2000
ER -