Key Techniques And Implementation Of Resource Joint Programming In Layered Transport Networks

With the rapid development of networking techniques and the multi-aspect growing of user requirements, the traditional networks are evolving into Next Generation Networks (NGN), and the architectures of transport networks are also changing gradually. To guarantee the end-to-end Quality of Service (QoS) and to enhance the utilization of overall network resources, researchers have studied the transport networks as layered or hierarchical structure, which is becoming one of the present hotspot issues of Next Generation Transport Network (NGTN). However, the QoS routing policies were generally addressed in only two layers network, some issues remain to be studied further, such as multi-layer network route selection, network resource optimization and income differences of different services.Sponsored by the technology fund project named "Key Techniques and Software Implementation of Large-scale Layered-network & Resource Joint Programming" of Huawei Technologies Co. Ltd, this dissertation presents a routing optimization algorithm, which uses a proposed object function and takes the characteristics in NGTN and its QoS requirements into account, and thus solving the drawbacks of published researchings. That algorithm is based on the Shortest Path First Algorithm, Dijkstra, which not only meets the QoS requirements of different users and differentiates incomes of diverse classes of services, but also achieves higher utilization of whole network resources and increases the total income of operators' services. In the dissertation, the architectures and network resource optimization algorithms of transport networks are studied in detail. The improvement on existing algorithms includes the proposed routing optimization object function and increasing the number of network optimization layers. The performance of the proposed schemes is analyzed and validated with computer simulation in the view of multi-aspects. Finally, the summary of whole dissertation is given and the issues to be studied further are also proposed.