This project contains a implementation of the algorithm for finding Steiner Minimal Trees in higher dimensions, as described by Smith in this article from Algorithmica, 1992.

The implementation tries to do some things smarter than the original implementation by Smith, e.g. by not rebuilding the topology every time a point is added or removed.

Furthermore the implementation also implements another iteration, apart from the one described by Smith. This iteration uses a analytical method to place Steiner points immediatly at the correct place.

The implementation also has the ability to sort points before optimizing. The sorting is an approximation of the longest path through a fully connected graph.

• golang 1.4 or newer

Simply run make and everything should be done.

Trees should be defined in separate JSON-files containing a key "points" having a list of all the points of the tree.

After building the program it can be used to optimize a tree defined in some file e.g. example.json by calling ./steinertree example.json

The program currently has the following flags which can be used:

• -1 : Change the index of edges and points in the output to start at 1 instead of 0. This is useful for comparing with the old implementation which is 1-indexed.
• -sort : Sort the terminals before optimizing.
• -iteration=[itername] : Select which iteration to use for optimization. Possiblities are: simple and smith. Default: smith