| The intelligent optical networks employing Routing and Wavelength Assignment (RWA) are considered as the most competitive candidates for next generation backbone networks. RWA is an important issue in intelligent networks, which refers to route finding and wavelength assignment when a lightpath is to be established between a node pair. A key issue in networks design is to optimize Routing and Wavelength Assignment (RWA) for lightpath. RWA solves how to find out an appropriate lightpath and assign a wavelength reasonably in order to make full use of the limited resource and provide communication capability as large as possible. The RWA includes static RWA and dynamic RWA. In this thesis, firstly, the basic conceptions and existing algorithms for RWA problem are presented, this thesis makes comparative research on the intelligent optimization algorithm for dynamic RWA.In order to resolve the problem of Routing and Wavelength Assignment (RWA) more efficiently, an algorithm in terms of contemporary routing is proposed. The algorithm based on the layered-graph model is used to resolve the two sub-problems of routing and wavelength assignment simultaneously. This paper studies the wavelength-routing all-optical networks without wavelength converter, the algorithm employs the layered-graph model, adjusts strategy of selecting routes dynamically from the limited number of wavelengths, attempts to efficiently utilizing resource of bandwidth under the wavelength-continuity constraint. Finally, OPNET is used to modeling and simulate. Simulation results show that blocking probability can be significantly reduced compared with First-Fit (FF) algorithm. Fairness is improved significantly too.At last, this paper studies network capability of the WDM all-optical networks with wavelength converter. Simulation results show that blocking probability can be significantly reduced after using wavelength converter. Consequently, the network integrated performance is improved. |
PDF Full Text Request