Lawler-Moore Speedups via Additive Combinatorics

TL;DR

This paper improves the Lawler-Moore dynamic programming framework for scheduling problems by introducing a new state-pruning method based on additive combinatorics, reducing dependence on total processing time to dependence on maximum processing time.

Contribution

The authors present a novel speedup of the Lawler-Moore recurrence using a state-pruning technique and an additive-combinatorial swapping argument, leading to faster algorithms for certain scheduling objectives.

Findings

Achieved algorithms with runtime depending on p_max^2 instead of P.

Provided the first near-linear-time algorithms for specific scheduling problems.

Improved the theoretical runtime bounds for classical scheduling objectives.

Abstract

The Lawler-Moore dynamic programming framework is a classical tool in scheduling on parallel machines. It applies when the objective is regular, i.e. monotone in job completion times, and each machine follows a fixed priority order such as Smith's Rule or Jackson's Rule. For the basic objectives $Pm||\sum w_jC_j$, $Pm||L_{\max}$, and $Pm||\sum w_jU_j$, it gives running times $O(P^{m-1}n)$, $O(P^{m-1}n)$, and $O(P^mn)$, respectively, where $P$ is the total processing time. Recent SETH-based lower bounds indicate that the dependence on $P$ is essentially optimal, but they do not rule out improved dependence on the maximum processing time $p_{\max}$. We give the first major speedup of the Lawler-Moore recurrence. Our main ingredients are a new state-pruning method and a swapping argument based on an additive-combinatorial lemma. We prove that, whenever this swap does not increase the objective value, there exists an optimal schedule in which, for every prefix of jobs, the load difference between any two machines is at most $4p_{\max}^2$. This lets us prune redundant states throughout the dynamic program, replacing the dependence on $P$ by a dependence on $p_{\max}^2$. We show that the swap is non-increasing for all three objectives above. Hence $Pm||\sum w_jC_j$ and $Pm||L_{\max}$ admit algorithms with running time $O(p_{\max}^{2m-2}n)$, while $Pm||\sum w_jU_j$ can be solved in time $O(p_{\max}^{2m-2}Pn)\le O(p_{\max}^{2m-1}n^2)$. These bounds strictly improve the original Lawler-Moore runtimes whenever $p_{\max}=o(\sqrt{P})$. In particular, for $Pm||\sum w_jC_j$ and $Pm||L_{\max}$, we obtain the first near-linear-time algorithms when processing times are polylogarithmic in $n$.