Link set capacity augmentation algorithms for networks supporting SMDS

To expand the link set capacities for networks supporting the switched multi-megabit data service (SMDS), one must determine how much additional capacity is needed and where it is needed so as to minimize the total capacity augmentation cost. The authors consider two combinatorial optimization problem formulations. These two formulations are compared for their relative applicability and complexity. Solution procedures based upon Lagrangean relaxation are proposed for the formulations. In computational experiments, it is clearly demonstrated that there is a computation-time versus solution-quality trade-off between the algorithms for the two formulations. In addition, to demonstrate the effectiveness of the proposed algorithms, the author compares the proposed algorithms with a most congested first (MCF) heuristic. For the test networks, the proposed algorithms achieve up to 250% (79% on the average) improvement in the total capacity augmentation cost over the MCF heuristic.<
>