| This paper is concerned with the minimum maximal flow problem, i.e., a problem of minimizing the flow value attained by a maximal flow for a given network. The optimal value indicates how inefficiently the network can be utilized under restricted controllability. We discuss the extension of the gap function defining the set of all maximal flows and then formulate the problem as a D.C.\ optimization problem. Based on this formulation, we propose the cut and split method combined with a local search technique. |
|