| Popis: |
Most of the studies on traffic grooming focus on minimizing network link capacity and providing serving-relationship between client services and link capacity. Subsequent to this step, it is important to plan for adding/dropping client services over client service ports and setting up end-to-end lightpaths over network ports, which is however seldom investigated. We call such effort node hardware module planning. This is an industrially practical problem aiming to minimize node hardware cost since hardware modules are the most expensive components in a network. Based on a link-based traffic grooming result that provides information on end-to-end capacity units incident to nodes and aggregation relationship between client services and capacity units, we develop an Integer Linear Programming (ILP) model to optimally plan hardware modules. To overcome the computation difficulty of the ILP model under large-sized planning scenarios, we also develop a fast sub-optimal heuristic for hardware module planning. Simulation studies indicate that the heuristic is efficient to achieve a design close to an optimal solution obtained by the ILP model. Also, the evaluation of the impact of switch backplane size shows that given a certain set of network modules, an optimal switch backplane size exists, which achieves the lowest hardware cost. |
