Online List Labeling: Breaking the $\log^2n$ Barrier
Abstract
The online list labeling problem is an algorithmic primitive with a large literature of upper bounds, lower bounds, and applications. The goal is to store a dynamically-changing set of items in an array of slots, while maintaining the invariant that the items appear in sorted order, and while minimizing the relabeling cost, defined to be the number of items that are moved per insertion/deletion. For the linear regime, where , an upper bound of on the relabeling cost has been known since 1981. A lower bound of is known for deterministic algorithms and for so-called smooth algorithms, but the best general lower bound remains . The central open question in the field is whether is optimal for all algorithms. In this paper, we give a randomized data structure that achieves an expected relabeling cost of per operation. More generally, if for , the expected relabeling cost becomes . Our solution is history independent, meaning that the state of the data structure is independent of the order in which items are inserted/deleted. For history-independent data structures, we also prove a matching lower bound: for all between and some sufficiently small positive constant, the optimal expected cost for history-independent list-labeling solutions is .
Cite
@article{arxiv.2203.02763,
title = {Online List Labeling: Breaking the $\log^2n$ Barrier},
author = {Michael A. Bender and Alex Conway and Martín Farach-Colton and Hanna Komlós and William Kuszmaul and Nicole Wein},
journal= {arXiv preprint arXiv:2203.02763},
year = {2022}
}
Comments
Full version for FOCS 2022 camera ready