Abstract:In this paper, we describe a method for solving some open problems in design theory based on SAT solvers. Modern SAT solvers are efficient and can produce unsatisfiability proofs. However, the stateof- the-art SAT solvers cannot solve the so-called large set problem of idempotent quasigroups. Two idempotent quasigroups over the same set of elements are said to be disjoint if at any position other than the main diagonal, the two elements from the two idempotent quasigroups at the same position are different. A collection of n-2 idempotent quasigroups of order n is called a large set if all idempotent quasigroups are mutually disjoint, denoted by LIQ(n). The existence of LIQ(n) satisfying certain identities has been a challenge for modern SAT solvers even if n = 9. We will use a finite-model generator to help the SAT solver avoiding symmetric search spaces, and take advantages of both first order logic and the SAT techniques. Furthermore, we use an incremental search strategy to find a maximum number of disjoint idempotent quasigroups, thus deciding the non-existence of large sets. The experimental results show that our method is highly efficient. The use of symmetry breaking is crucial to allow us to solve some instances in reasonable time.

Investigating the Existence of Costas Latin Squares via Satisfiability Testing

Investigating the Existence of Large Sets of Idempotent Quasigroups via Satisfiability Testing.

Investigating the Existence of Orthogonal Golf Designs via Satisfiability Testing

Fun with Latin Squares

Completing partial Latin squares with two filled rows and three filled columns

Do K33-Free Latin Squares Exist?

Distance in Latin Squares

Latin squares without proper subsquares

On the Number of Maximum Inner Distance Latin Squares

An Efficient Local Search for Partial Latin Square Extension Problem

Constructing Orthogonal Latin Squares from Linear Cellular Automata

Latin squares with five disjoint subsquares

On completing three cyclic transversals to a latin square

Limits of Latin squares

Subsquares in random Latin squares and rectangles

Unsatisfiability Proofs for Weight 16 Codewords in Lam's Problem

Incomplete MaxSAT Approaches for Combinatorial Testing

Random Latin squares and Sudoku designs generation

Some Relations on Paratopisms and an Intuitive Interpretation on the Adjugates of a Latin Square

Enumeration of Latin squares with conjugate symmetry

Canonical labelling of Latin squares in average-case polynomial time