A new metaheuristic and its application to the Steiner problems in graphs

Metaheuristics are a set of related ideas that provide a general framework to obtain the approximated solution of combinatorial optimization problems. We present ideas for a new method that we call SN. SN is based on the idea of dividing a given optimization problem into decision subproblems which are easier to solve heuristically. In order to verify the effectiveness of this metaheuristic approach, we developed an algorithm to solve a classic graph theory problem, the Steiner Problem in Graphs. Preliminary results show that the new algorithm is competitive when compared with the best heuristic algorithms known for this problem.