Abstract
Determining the maximal density $m_1(\mathbb{R}^2)$ of planar sets withoutunit distances is a fundamental problem in combinatorial geometry. This paperinvestigates lower bounds for this quantity. We introduce a novel approach toestimating $m_1(\mathbb{R}^2)$ by reformulating the problem as a MaximalIndependent Set (MIS) problem on graphs constructed from flat torus, focusingon periodic sets with respect to two non-collinear vectors. Our experimentalresults supported by theoretical justifications of proposed method demonstratethat for a sufficiently wide range of parameters this approach does not improvethe known lower bound $0.22936 \le m_1(\mathbb{R}^2)$. The best discrete setsfound are approximations of Croft's construction. In addition, several opensource software packages for MIS problem are compared on this task.