Study on Planning and Control in Translucent Wavelength Switched Optical Network with Physical Layer Impairment Constraint

Wavelength division multiplexing has been considered as one of the promising technologies for the future broadband networks. In order to take advantage of the potential of optical transmission in long distance and huge bandwidth in a more cost-efficient way, the concept of wavelength switched end-to-end lightpath has been introduced into optical networks. Such a network is termed as the wavelength switched optical network (WSON). WSON is expected to provision lightpaths for upper-layer transport networks, or, offer dynamic on-demand lightpath services for bandwidth-consuming applications, such as e-business, e-science, online games, etc. In general, the WSON can be modelled as the interaction of three functional planes: data plane, control plane, and management plane. In the data plane, the blocking performance in WSON is strongly subject to the wavelength continuity constraint and wavelength exclusivity constraint. The wavelength converter can be used to mitigate the effect of the first constraint. Due to the physical layer impairment accumulated along the route traversed by the optical signal, the optical signal will be deformed to be unrecognizable at the destination if it is not regenerated after a certain distance transmission. The degraded optical signal can be reformed with reamplifying, reshaping, and retiming, namely 3R regeneration. At present, the optical-electronic-optical (O/E/O) technology is the most reliable and mature method to realize these functions, which gives rise to the translucent WSONs. In literatures, a great deal of research has been carried out on the placement of wavelength converters or 3R regenerators separately. Meanwhile, in many cases, based on optoelectronic conversion, these functions can be technically integrated into a single shared physical component, namely the wavelength convertible 3R regenerator (WC3R). It will be more efficient to use this multi-function component compared to using wavelength converter and 3R regenerator individually. This serves as a motivation behind the investigation of the sparse placement issue of WC3Rs presented in this dissertation, which also distinguishes the network design problem in the translucent WSON from that in the conventional transparent WSON. Given a carefully designed translucent WSON, the control functions need to be conceived to provision lightpaths through the network in an efficient way. The concept of these functions mainly involves routing and resource assignment (RRA). As the lack of flexibility together with the scalability incapacitates the centralized approach for future, my study here in the control plane is emphasized in the distributed system. Recently, the application of generalized multi-protocol label switching (GMPLS) protocol suite …