2 ALGORITHMS FOR THE SUBSET INTERCONNECTION DESIGN PROBLEM

An NP-complete generalization of the minimum spanning tree problem is considered. Given are a set of V, a cost function c: V x V --> R+, and a collection {X1 ,..., X(m)} of subsets of V. A graph G with vertex set V is called feasible if every X(i) induces a connected subgraph of G. The minimum subset interconnection design problem is to find a feasible graph with a minimum cost sum. In this paper, two heuristic algorithms for the problem are given and analyzed. Several classes of input data for which one of the algorithms finds optimal or at least approximative solutions are presented.