Integrated route selection, transponder placement, wavelength assignment, and restoration in an advanced ROADM architecture

New reconfigurable optical add-drop (ROADM) architectures based on rapidly tunable transponders and wavelength selective switches are being actively studied in the Defense Advanced Research Projects Agency (DARPA) core optical networks project and elsewhere. They are making new optical network architectures possible, and could change optical network cost structures. This paper focuses on two principal issues arising from these new designs: (i) What should be the relationships between the routing, wavelength assignment, transponder placement, and restoration planning functions for these networks? (ii) How should the new cost structures affect their design? We show for a representative large core network employing shared mesh restoration to offer dynamic wavelength services that integrating these functions and dealing with trade-offs between them using a cost-based methodology can significantly reduce total network cost when compared with non-integrated methods or methods that focus on optimizing specific physical metrics such as wavelength mileage or regenerator counts.