| Abstract: | With the progress of science and technology, human’s demands for the telecommunication services are more and more diversified. From the traditional voice communication to huge volume of data transmission, the telecommunication networks have already passed many decades and have a new style. Today, our customers demand high quality of communication, not only requiring correct transmission, short delays, but also sufficient control over the bandwidth. Most new-generation networking systems adopt the Multi-Protocol Label Switching (MPLS) protocol as the core technology for transmission. So far, its superior transmission quality has proved to meet a lot of requirements. Hence, following the similar ways of operation, the telecommunication network providers develop the GMPLS（Generalized Multi-Protocol Label Switching）technology.
Expanding from the MPLS technology, GMLPS can switch between various transmission modes (e.g., Packet Switching, Layer 2 Switching, Time Division Multiplexing, Lambda Switching, Fiber Switching) for providing different levels of applications. Hence, GMPLS are qualified enough to be the standard of future network applications. The transmission modes can be roughly classified into two categories： Telecom and Datacom. They are greatly different from each other, especially at the control plane. GMPLS has unified the control plane, which makes a common platform to manage for different patterns of switching. Based on this characteristic, the line state database（Link State Database）and the flow project line database（TE Link Database）can be managed together. Current research topics in GMPLS are mostly on discussing the fairness, reliability and efficiency of the network policy, such as the route distribution, flow engineering of the network. All these viewpoints come from network efficiency without considering the network operations of the operators.
This thesis puts forward a new way of thinking about the network operations, not merely focusing on the transmission efficiency of the networks. We propose a weighted route distribution method to assign the routes of a network. Fuzzy theory is used to assign imprecise values and different weights to various routes based on their reliability and service classes. While distributing the protection routes, we assign important transmission services on the routes of high weights（high dependability）. By utilizing this route distribution method, we can differentiate the service quality, reduce the risk of the high-value products, and promote the satisfaction level of our customers. |
