Research On Resource Allocation Model And New Optimization Algorithms In Elastic Optical Networks With Multi-core Fibers Title Of The Thesis For Professional Master's

With the rapid development of Internet technology and its applications,the network bandwidth requirement is increasing explosively.The provision capabilities of large capacity transport and dynamic flexible configuration become the trend of future networks development.For the increasing demand of transmission capacity,both the wavelength division multiplexing networks used currently and Elastic Optical Networks(EONs)can not effectively solve the problem that the transmission capacity of an existing standard single-core fiber is approaching to its Shannon limit.In recent years,a new EON based on multi-core fiber Space Division Multiplexing(SDM)technology has been proposed.This optical network can achieve a significant increase in network capacity,which can well meet the continuous needs of the expansion traffic.Resource allocation algorithm of EONs with multi-core fibers has a great impact on the usage of network resources,and it is more difficult compared with single-core fiber network,which makes it an important research issue.Sponsored by the project from the National Science Foundation of China(NSFC),the novel intelligent algorithms and formalization of resource optimization in hybrid EONs,this thesis studies the resource optimization problem of the static traffic in EONs with multi-core fibers.This thesis briefly introduces the development status of optical network technology and the research status of EONs with the optical fiber of new types.The composition,the main components,the key technologies of EONs and the research status of multi-core fiber are summarized.We discuss the routing,spectrum and core allocation(RSCA)problem in EONs with multi-core fibers and its research status.In addition,the relevant theoretical knowledge of multi-objective evolutionary algorithm is briefly described.The main contributions of this dissertation include the following two parts:(1)Resulted from the crosstalk among adjacent cores of a multi-core fiber,there exists a larger challenge for the resource allocation in EONs with multi-core fibers.The analysis on RSCA problem in EONs with multi-core fibers is made in this thesis.We establish a resource optimization model that considers the crosstalk among adjacent cores and minimizes the maximum sequence number of frequency slots occupied.To the best of our knowledge,an algorithm based on co-evolution is proposed to solve the RSCA problem forthe first time.In the proposed algorithm,the Most Service First(MSF)strategy is used to sort the traffic demand in descending order,and then the different initial populations with integer coding are set up for the routing sub-population and core allocation sub-population,respectively.Combined with single point crossover and uniform mutation,the global optimal solution of the RSCA problem is searched by the differentiated elite reserve evolution strategy.The performance evaluation software of the proposed algorithm is implemented via MATLAB.To verify the effectiveness of the proposed algorithm,the numerous simulation experiments are carried out on two network topologies of NSFNET and CHNNET.Experimental results show that the proposed algorithm could realize the efficient use of the network spectrum resources and make the traffic bearing distribution more uniform.(2)From the view of network overall operation,it is necessary to pay attention to the influence of the total crosstalk on traffic transmission in EONs with multi-core fibers.Considering EONs with multi-core fibers with limited resources,we assign the crosstalk of the whole network as one of the optimization objectives.A bi-objective optimization model is proposed to minimize both the blocking probability and the crosstalk among different cores simultaneously.To solve this optimization model,we propose a bi-objective genetic algorithm that is based on traditional decomposition strategy and uniform design.The proposed algorithm firstly generates the initial population using uniform design,and then set evenly distributed weight vectors to decompose the bi-objective optimization problem into a series of single-objective optimization sub-problems on the basis of Tchebycheff decomposition.Finally,simulation experiments are carried out in two general network topologies,and experimental results indicate the effectiveness of the proposed algorithm.