Research On Resource Allocation Methods Based On Crosstalk Aware In Space Division Multiplexing Elastic Optical Networks

With the growth of emerging services,the network bandwidth increases exponentially.The elastic optical networks(EONs)based on optical orthogonal spectrum multiplexing have fine spectrum granularity,which can flexibly allocate spectrum resources according to the services bandwidth.And it has became the next generation optical networks with great development potential.However,with the growth of traffic,EONs based on single-core fiber has been unable to meet the high-capacity services requirements.The space division multiplexing(SDM)technology using multi-core fiber,which can expand the transmission capacity of optical networks from the physical structure.The space division multiplexing elastic optical networks(SDM-EONs)considers the advantage of EONs and SDM,which can solve the problem of insufficient bandwidth of big data transmission,but it increases new challenges.Firstly,when allocating spectrum resources in SDM-EONs,three constraints of spectrum allocation must be observed,which greatly limits the flexibility of spectrum allocation,and makes the spectrum resources in the networks more discrete.Secondly,the multi-core fiber increases the physical capacity,but brings the problem of inter-core crosstalk,which affects the transmission quality of network services.Based on the analysis above,it is important to design resource allocation algorithm to optimize the inter-core crosstalk and spectrum fragmentation.This paper will mainly study the resource allocation algorithm based on inter-core crosstalk and spectrum fragmentation in SDM-EONs.The main work and contributions are as follows:Considering the resource allocation in SDM-EONs,chapter 3 considers spatial and frequency domain resources and proposes a resource allocation algorithm based on the importance of optical nodes and crosstalk aware.First of all,according to the importance of the optical nodes,the LRSC is sparsely configured in the optical networks.Secondly,in the path selection stage,according to the shortest path algorithm,k candidate paths are found.Then,according to the path load and the usage of the LRSCs in the path,the path weight formula is designed,and the candidate paths are sorted in ascending order according to the calculated weight values.In the spectrum allocation stage,the cores are grouped based on vertex coloring.Finally,according to the size and proportion of services,the spectrum is partitioned.Then,a core spectrum selection scheme considering inter-core crosstalk and core load is designed,which reduces the blocking probability of services.The simulation results show that compared with the comparison algorithm,the proposed algorithm can improve the network performance,its bandwidth blocking probability is optimized to 9% and spectrum utilization is up to 8%.Based on chapter 3,the chapter 4 considers the resource allocation in frequency domain,time domain and space domain.In order to reduce the inter-core crosstalk and fragmentation,a resource allocation algorithm considering holding time,the neighborhood matching degree and inter-core crosstalk is proposed.Firstly,in the routing stage,k candidate paths are found according to the shortest path algorithm.And k alternative paths are descending order according to the path load and path hops.Secondly,in order to reduce the spectrum fragmentation and inter-core crosstalk in the process of service transmission.In the core selection stage,the weight formula of core selection is designed,which comprehensively considers the load of core and the crosstalk between adjacent cores.In the spectrum allocation stage,according to the change of the time-frequency fragment and neighborhood matching degree,the spectrum block is selected,which aims at achieving load balance and reducing the increase of spectrum fragment.Finally,for the service that can not be successfully transmitted,the spectrum segmentation strategy will be adopted to divide the services into two small bandwidth services.And the appropriate spectrum resource is selected in the path for the services.Simulation results show that the proposed algorithm can effectively reduce spectrum fragmentation and improve bandwidth blocking probability and spectrum utilization,its bandwidth blocking probability is optimized to 9% and spectrum utilization is up to 6%.