Diverse Routing of Scheduled Lightpath Demands
ABSTRACT This article addresses the problem of defining working and protection paths for Scheduled Lightpath Demands (SLDs) in an optical transport network. An SLD is a demand for a set of lightpaths (connections), defined by a tuple (s, d, n, #, #), where s and d are the source and destination nodes of the lightpaths, n is the number of requested lightpaths and #, # are the set-up and tear-down dates of the lightpaths. The problem is formulated as a combinatorial optimization problem where the objective is to minimize the number of channels required to instantiate the lightpaths. Two techniques are used to achieve this goal: channel reuse and backup-multiplexing. The former consists of assigning the same channel (either working or spare) to several lightpaths, provided that these lightpaths are not simultaneous in time. The latter consists of sharing a spare channel among multiple lightpaths. A spare channel cannot be shared if two conditions hold: a) the working paths of these lightpaths have at least one span in common and b) these lightpaths are simultaneous in time. In the other cases, the spare channel can be shared. We propose a Simulated Annealing (SA) based algorithm to find approximate solutions to this optimization problem since finding exact solutions is computationally intractable. The results show that backup-multiplexing improves the utilization of channels but requires significant computing capacity. Under a fixed computing capacity budget, the technique is useful in cases where there is little time disjointness among SLDs.