Study On Wavelength Routing And Optimal Architecture In Intelligent Optical Networks

Wavelength routing optical network (WRON) is a practicable technology for next-generation high-speed backbone networks. Due to the limitation of wavelength resource, it is an important problem in WRON that how to utilize limited wavelength resources and improve the blocking probability of the networks by improving the capacity of the networks. There are two methods to solve the above problem. One is by using optimal routing and wavelength assignment (RWA), the other is by using optimal configuration of wavelength conversion. This thesis contains wide research on optimal routing and wavelength assignment and optimal placement of wavelength converters. The main parts are as follows:(1) Based on the learning of references, the importance and the effect to WDM of RWA and wavelength conversion were discussed. The algorithms, analytic models, and results obtained by other researchers were summarized.(2) We have deeply studied RWA problem in all-optical networks and submitted novel original algorithms of static and dynamic RWA. Compared with normal algorithms, we found that the performance of our algorithms was very excellent and was better than that of optimal algorithms submitted by others. By eliminating the lonely island and reducing the degree of some nodes easily blocked, novel static algorithm balanced graph connectivity to improve algorithm performance. Based on traffic load in networks, novel centralized dynamic algorithm dynamically adjusted path cost and revised routing path table to reduce blocking probability. By introducing pre-releasing scheme and overtime-releasing scheme, novel distributed dynamic algorithm could effectively release wavelength resource.(3) Based on previous models, we extended that and derived an analytical expression to compute the blocking probabilities in optical networks. Then we presented our optimal converter placement algorithm based on genetic algorithm to solve this NP-C problem and analyzed relationship between wavelength conversion and RWA. By adjusting mutation value and crossover value, novel genetic algorithm formed a positive feedback to accelerate searching speed and convergence speed.(4) By using optical switches, we proposed a new kind of wavelength router architecture. It overcame the wavelength continuity constraint and greatly increased ratio of wavelength reuse and decreased the blocking probability by placing wavelength converters. Simultaneously, we designed our wavelength router control system and developed router software control module. We realized a whole wavelength router system.