Competitiveness of the mSTreg heuristic beyond BCST and CST

Determine whether the mSTreg heuristic, which alternates geometric optimization of Steiner point positions with topology updates via minimum spanning trees, achieves competitive performance on other tree optimization problems—specifically including the general optimum communication spanning tree—where edge costs are weighted by different functions.

Background

The mSTreg heuristic is proposed to approximate BCST (and CST) by iterating between Steiner point geometry optimization and recomputation of an MST over terminals and Steiner points. The method is agnostic to the specific weighting of edge lengths, suggesting potential applicability to other problems such as the optimum communication spanning tree.

The authors explicitly highlight the need to test whether this heuristic remains competitive beyond the settings studied in the paper.