The Optical Network Optimization Software Design And Key Technologies

Recent few years, with the explosive growth of network traffic and great development of network technology, all operators need to continually optimize present network in order to enhance the competition of them. How to scientifically optimize network has become a focus. Making full use of network optimizing software can bring great efficiency and veracity. Due to most present network optimizing software are applied to wireless network, while few of them are designed for optical network, exploiting an optical network optimizing software system has huge market value.Considering the demand of a certain great corporation, the first research work of this thesis focuses on the study and development of optical network optimizing software system, and its whole design process strictly follows the engineering design rules. After completing the requirement analysis and the frame design of the software system, the technologies of network topology optimizing are studied, for the purpose of accomplishing topology optimizing function which is one of core functions in the system. Based on the analysis and compare of various methods of optimizing two typical topologies, ring-like and mesh-like, it respectively develops improved methods. At the same time, considering the difficulty of optimizing the network with huge size and complicated framework, the region partition strategy is studied. And through the study, a feasible region partition algorithm and a flow of correlative operation after region partition are proposed in this thesis. This region partition algorithm can effectively reduce the complexity of the optimizing of huge network, and the results of strategy are also practical. With these theoretical researches, this thesis gives the detailed design of the topology optimizing module in the software system, and validates the usability of its each function through cell tests.For the importance of Survivability in network optimizing, this thesis also focuses on the dynamic traffic distribution problem under shared risk link group (SRLG) constraints, and a new improved algorithm is proposed. This algorithm can effectively avoid traps in path protection under SRLG constraints and it introduces resource mixed shared scheme for ensuring high resource utilization. Differing from previous shared path protection scheme, this scheme make the primary path, when it doesn't work, share its resource with the backup paths. Simulation results show that the resource utilization and effective blocking probability of improved algorithm are better than shared path protection scheme.To evaluate the performance of the algorithm presented in this thesis, a simulator is developed with the principle of C++ Object Oriented Design. The main frame of the software and the detail design of each module are given in Chapter 4.