Constraint-based local search for Golomb rulers

Abstract

This paper presents a constraint-based local search algorithm to find an optimal Golomb ruler of a specified order. While the state-of-the-art search algorithms for Golomb rulers hybridise a range of sophisticated techniques, our algorithm relies on simple tabu meta-heuristics and constraint-driven variable selection heuristics. Given a reasonable time limit, our algorithm effectively finds 16-mark optimal rulers with success rate 60 % and 17-mark rulers with 6 % near-optimality.