| With the increasing of international Internet traffic,undersea cable systems have developed rapidly.The task of the future submarine cable is not only communication and transmission but also information detection and processing.How to satisfy the needs of transmitting multiple requests in one network simultaneously.It is a major issue for scholars in the world.The research in this thesis is dedicated to future underwater optical networks.We have designed future underwater optical networks at different levels,ranging from physical route,routing,and wavelength assignment.The main contents are as follows:Firstly,this paper introduces the development status of underwater optical networks.The reason why the structure of the underwater optical network cannot meet the needs of the future network is analyzed.The physical route planning evaluation model is designed.By comparing the simulation results,the author select a layered design scheme for the future underwater optical network with moderate cost,good network performance,high reliability and easy expansion.Secondly,this paper points out that the traditional routing algorithm can cause a load imbalance problem on the planning scheme topology.The author combined the advantages of fixed routing algorithm and adaptive routing algorithm and designed a load balance routing algorithm in order to solve the problem.Thirdly,there is a problem that massive data is transmitted quickly,efficiently,flexibly and accurately in the future underwater optical networks.The author proposes a wavelength assignment strategy based on the priority of the request in order to solve the problem.According to this assignment strategy,the wavelength assignment algorithm based on the priority of the request is designed.Lastly,the authors combine the load-balancing routing algorithm with the wavelength assignment algorithm,and request adaptive routing algorithm applied to future underwater optical network.The simulation results show that compared with the traditional routing and wavelength assignment algorithms,this algorithm performs better in terms of load balancing,network blocking rate,service failure rate and link utilization.At the same time,the feasibility of assignment strategy is demonstrated by comparing simulation results with different priorities. |