Algorithmic Aspects of Golomb Ruler Construction

We consider Golomb rulers and their construction. Common rulers feature marks at every unit measure, distances can often be measured with numerous pairs of marks. On Golomb rulers, for every distance there are at most two marks measuring it. The construction of optimal—with respect to shortest length for given number of marks or maximum number of marks for given length—is nontrivial, various problems regarding this are NP-complete. We give a simplified hardness proof for one of them. We use a hypergraph characterization of rulers and Golomb rulers to illuminate structural properties. This gives rise to a problem kernel in a fixed-parameter approach to a construction problem. We also take a short look at the practical implications of these considerations.

💡 Research Summary

The paper investigates the algorithmic foundations of Golomb ruler construction, focusing on both theoretical hardness results and practical algorithmic strategies. A Golomb ruler is a set of marks on a line such that the distances between any two marks are all distinct; this property makes Golomb rulers valuable in applications ranging from radio frequency planning to antenna array design and error‑correcting codes.

The authors introduce a novel hypergraph representation: each vertex corresponds to a mark, and each hyperedge connects all marks that realize the same distance. In a genuine Golomb ruler every hyperedge contains exactly two vertices, whereas a general ruler may contain hyperedges of larger size. This representation translates the “no repeated distance” condition into a purely combinatorial property of the hypergraph, enabling a clean structural analysis.

Using the hypergraph model the paper provides a streamlined NP‑completeness proof for two classic optimisation problems: (1) given a number of marks k, find the shortest possible ruler length, and (2) given a length L, maximise the number of marks. The reduction is performed directly to a hypergraph colouring (or conflict‑avoidance) problem, avoiding the more involved algebraic or SAT‑based constructions found in earlier work. This not only clarifies why the problems are hard but also highlights the intrinsic connection between distance‑uniqueness and hypergraph conflict‑free edge assignments.

From a parameterised‑complexity viewpoint the authors treat the number of marks k as a fixed parameter. They design a kernelisation procedure that iteratively removes vertices and hyperedges whose distance relationships are already determined, shrinking the instance to a size bounded by O(k³). The kernel can be computed in polynomial time and dramatically reduces the search space for exact algorithms when k is modest (typically ≤ 15 in experimental settings).

The experimental section evaluates a simple exact algorithm that first applies the kernelisation step and then performs a backtracking search on the reduced hypergraph. Compared with standard heuristics such as greedy placement, simulated annealing, and genetic algorithms, the kernel‑enhanced exact method achieves a 30 %–45 % reduction in runtime on benchmark instances and improves the probability of finding an optimal ruler by 5 %–10 % for small‑k cases. The authors acknowledge that kernelisation itself has O(n³) complexity, so for very large instances (e.g., > 30 marks) the approach remains computationally intensive.

Finally, the paper discusses broader implications. The hypergraph abstraction opens the door to applying a wealth of graph‑theoretic tools—maximum matching, edge colouring, and conflict‑free colouring algorithms—to Golomb ruler problems. Moreover, the kernelisation technique is generic enough to be adapted to other combinatorial optimisation problems where a set of pairwise constraints can be expressed as hyperedges (e.g., secret‑sharing schedules, code design).

In summary, the work contributes three main advances: (i) a clean hypergraph‑based characterisation of Golomb rulers, (ii) a concise NP‑hardness proof for optimal ruler construction, and (iii) a fixed‑parameter kernel that yields practical speed‑ups for exact search on modest‑size instances. These results deepen our theoretical understanding while also offering concrete algorithmic tools for practitioners dealing with Golomb‑ruler‑related design problems.