All-optical Network Topology Design And Wavelength Allocation Problem

The information society is now changing many aspects of our life. In this society, the demand and dependence on the information are the main characters. With the coming of the INTERNET era, a lot of new information service will be in use and the need for the bandwidth will become more and more. At the same time with the development of the technology of optical networks, the request for Quality of Service (QoS) is more and more urgent. In optical networks in particular and in networks with guaranteed quality of service in general, resource management is an important issue. Resources must be reserved for every established connection, and a bad resource management strategy may lead to reduced utilization and throughput of the network. However, optical network based on optical wavelength has its intrinsic defects. These cause the mismatch between the need and the traditional backbone network in which the SONET/SDH technology is dominant. It is urgent of the service provider to find another technology that has more performance-to-price than ever to overcome this mismatch. The WDM (Wavelength Division Multiplexing) technology is now coming on the stage to satisfy the urgent need. Because the WDM technology can fully utilize the enormous bandwidth of the optical fiber, the cost for transfering data can be greatly decreased. This enables WDM technology to be the most attractable technology in wide backbone network. It is predicted that in the near future, the WDM technology will be widely used in other networks.Firstly, the number of wavelengths that can be multiplexed in the same fiber is limited, which leads to a few numbers of channels identified by wavelength established in wavelength-routed network. Secondly, optical nodes usually do not possess the ability of wavelength conversion for the restriction of cost and technology. Optical network brings out so-called wavelength continuity constraint because connection must be established on the path that be carried by the same wavelength between source and destination nodes. In WDM networks, wavelength conversion is a key technology, but it is always the focus of the contention. Although wavelength conversion can improve the performance of WDM networks, simplify the control, and so on, it is difficult to manufacture the ideal all-optical wavelength converter due to the limit of the technology. And more the technology is now in the laboratory, when it is on commercial use, the price will be very expensive. So it is necessary to construct communication networks and manage resources efficiently.In all-optical networks, the number of channels (wavelengths) is a scarce resource. To make as efficient use as possible of the optical bandwidth, it is necessary to control two parameters: the number of paths that use the same link and the sum of required wavelengths. In this thesis, our goal is to establish a given set of connection requests with a minimum number of wavelengths. In graph-theoretic terms, the wavelength assignment problem can be viewed as assigning colors to paths so that paths sharing an edge receive different colors.A network topology structure can be indicated by a graph. We can study the structure of network by studying the character of the graph. The theory of studying the character of graphs is the graph theory and the coloring of graphs is one of the important contents. Generally, the coloring of graph is divided into vertex-coloring and edge-coloring. Edge-coloring is divided into proper edge-coloring and f-coloring. The colorings of graphs have applications in many fields, such as wavelength assignment in AON. Regular topologies arc good candidates as logical topologies since they are well understood and results regarding bounds and averages (e.g., for hop lengths) arc easier to derive. Algorithms for routing traffic on a regular topology are usually simple. Also, regular topologies possess inherent load balancing characteristics which are important when the objective is to minimize the maximum congestion. So we only consider regular topologies.We mainly study wavelength assignment problems in static all-optical networks in this thesis, and obtain some interesting results. The thesis is divided into five chapters.We illuminate the background of our research and propose the problems, the work we do and the structure of the thesis in the first chapter. Some preliminaries in graphs are given in the second chapter. In the third chapter we mainly study the logical topologies for wavelength-routed optical networks. RWA in wavelength-routed all optical networks arc studied in the fourth chapter. The fifth chapter mainly presents unsolved problems for future research.