We present an evolutionary programming algorithm for solving the dynamicrouting and wavelength assignment (DRWA) problem in optical wavelength-divisionmultiplexing (WDM) networks under wavelength continuity constraint. We assumean ideal physical channel and therefore neglect the blocking of connectionrequests due to the physical impairments. The problem formulation includessuitable constraints that enable the algorithm to balance the load among theindividuals and thus results in a lower blocking probability and lower meanexecution time than the existing bio-inspired algorithms available in theliterature for the DRWA problems. Three types of wavelength assignmenttechniques, such as First fit, Random, and Round Robin wavelength assignmenttechniques have been investigated here. The ability to guarantee both lowblocking probability without any wavelength converters and small delay makesthe improved algorithm very attractive for current optical switching networks.
展开▼
