Sparse Traffic Grooming in Translucent Optical Networks

Abstract

Traffic grooming is important in the design and operation of translucent optical networks. This paper investigates sub-wavelength traffic grooming in a translucent optical network and shows how to take advantage of sub-wavelength traffic grooming capability of sparsely distributed opaque switch nodes. An efficient heuristic, called the virtual nodal degree ranked algorithm, is employed to select the best locations for opaque switch nodes. New mixed-integer linear programming (MILP) optimization models are developed to optimally groom sub-wavelength traffic demand in the translucent optical network. The models maximize served sub-wavelength traffic demand under a limited network capacity and minimize required wavelength capacity subject to the condition that all the traffic demands are served. The models are novel in optimization methodology because they incorporate both the arc-node and the arc-path multi-commodity optimization techniques for a single problem. Based on simulation studies, it is found that the performance of sub-wavelength traffic grooming of a translucent network is substantially improved by increasing the number of opaque switch nodes. For some network topologies, the performance saturates as the number of opaque switch nodes increases, though the phenomenon is not general for any type of network topology.