Tight running times for minimum $\ell_q$-norm load balancing: beyond exponential dependencies on $1/\epsilon$

TL;DR

This paper presents a PTAS for the minimum ll_q-norm load balancing problem with a running time of 2^{ ilde{O}(\u221a{1/\u03b5})}+n^{O(1)}, surpassing the typical exponential dependency on 1/, and proves this is optimal under ETH.

Contribution

It introduces a novel PTAS with sub-exponential dependence on 1/ for a strongly NP-hard scheduling problem and establishes matching lower bounds under ETH.

Findings

The PTAS runs in 2^{e}()^{ ilde{O}(e)}+n^{O(1)} time.

The lower bound matches the upper bound under ETH, showing optimality.

Algorithms with better running times exist for specific ranges of machine count m.

Abstract

We consider a classical scheduling problem on $m$ identical machines. For an arbitrary constant $q>1$, the aim is to assign jobs to machines such that $\sum_{i=1}^m C_i^q$ is minimized, where $C_i$ is the total processing time of jobs assigned to machine $i$. It is well known that this problem is strongly NP-hard. Under mild assumptions, the running time of an $(1+\epsilon)$-approximation algorithm for a strongly NP-hard problem cannot be polynomial on $1/\epsilon$, unless $\text{P}=\text{NP}$. For most problems in the literature, this translates into algorithms with running time at least as large as $2^{\Omega(1/\varepsilon)}+n^{O(1)}$. For the natural scheduling problem above, we establish the existence of an algorithm which violates this threshold. More precisely, we design a PTAS that runs in $2^{\tilde{O}(\sqrt{1/\epsilon})}+n^{O(1)}$ time. This result is in sharp contrast to the closely related minimum makespan variant, where an exponential lower bound is known under the exponential time hypothesis (ETH). We complement our result with an essentially matching lower bound on the running time, showing that our algorithm is best-possible under ETH. The lower bound proof exploits new number-theoretical constructions for variants of progression-free sets, which might be of independent interest. Furthermore, we provide a fine-grained characterization on the running time of a PTAS for this problem depending on the relation between $\epsilon$ and the number of machines $m$. More precisely, our lower bound only holds when $m=\Theta(\sqrt{1/\epsilon})$. Better algorithms, that go beyond the lower bound, exist for other values of $m$. In particular, there even exists an algorithm with running time polynomial in $1/\epsilon$ if we restrict ourselves to instances with $m=\Omega(1/\epsilon\log^21/\epsilon)$.