Approximating the bottleneck plane perfect matching of a point set (Q904112)

\textit{A. K. Abu-Affash} et al. [ibid. 47, No. 3, Part A, 447--457 (2014; Zbl 1281.65025)] showed that the bottleneck plane perfect matching problem is NP-hard and presented an algorithm that computes a plane perfect matching whose edges have length at most \(2\sqrt {10}\) times the bottleneck. The present authors present an algorithm which computes a plane matching of size \(n/6\) in unit disk graphs (UDG) and observe that the problem of showing if this algorithm terminates in polynomial number of steps or not, is still open. Presenting a \(2/5\)-approximation algorithm for computing a maximum matching in UDG, and by extending it, the authors show how one can compute a bottleneck plane matching of size \(n/5\) with edges of optimum-length.