English

Load balancing under $d$-thinning

Probability 2020-01-06 v2

Abstract

In the classical balls-and-bins model, mm balls are allocated into nn bins one by one uniformly at random. In this note, we consider the dd-thinning variant of this model, in which the process is regulated in an on-line fashion as follows. For each ball, after a random bin has been selected, an overseer may decide, based on all previous history, whether to accept this bin or not. However, one of every dd consecutive suggested bins must be accepted. The maximum load of this setting is the number of balls in the most loaded bin. We show that after Θ(n)\Theta(n) balls have been allocated, the least maximum load achievable with high probability is (d+o(1))dlognloglognd(d+o(1))\sqrt[d]{\frac{d\log n}{\log\log n}}. This should be compared with the related dd-choice setting, in which the optimal maximum load achievable with high probability is loglognlogd+O(1)\frac{\log\log n}{\log d}+O(1).

Keywords

Cite

@article{arxiv.1908.10278,
  title  = {Load balancing under $d$-thinning},
  author = {Ohad N. Feldheim and Jiange Li},
  journal= {arXiv preprint arXiv:1908.10278},
  year   = {2020}
}

Comments

Title changed from "The power of $d$-thinning in load balancing"; minor change of abstract; add more backgrounds on thinning in introduction