In this paper we report computational experience with a branch and cut solver for the Steiner tree problem on a graph. The problem instances include complete graphs, randomly generated sparse graphs and grid graphs. The edge weights are either randomly generated or are the Euclidean distance between the endnodes that are placed at random on the plane. The effect of changing various problem parameters on solution time is studied.
|Original language||English (US)|
|Number of pages||16|
|Journal||ORSA journal on computing|
|State||Published - Jan 1 1992|
ASJC Scopus subject areas