Spectrum Reconstruction In Software Defined Multi-domain Optical Network

With the development of Internet data services,in response to the features of data services such as dynamic,emergent,elastic optical network technologies have emerged,that leads to routing and spectrum allocation problem(RSA).With the growth of the optical network,RSA issues in large scale multi-domain optical network become even more prominent.Software defined network(SDN)technique based on general computing resources which provides the control plane of programmability makes it possible to implement a fine defragmentation algorithm which based on abstract of the network.This paper inquires SDN in the scenario of multi-domain optical network,on this basis,study how to use the capacity of control and management provided by SDN to implement more intelligent and efficient spectrum reconstruction for multi-domain optical network.This paper's work and innovations including:1)Presents a hierarchical software defined multi-domain optical network control technology,build a prototype which is based on opendaylight and do function test in an experimental network to verify the feasibility of the control technology.2)Based on Utilization Entropy indicator and its statistics feature,design a mechanism to monitor spectrum status of the network and trigger defragmentation in multi-domain optical network.Propose a defragmentation algorithm based on k-shortest path algorithm and principle of minimum Utilization Entropy spectrum.Do simulation and experimental verification on a testbed to validate the validity of the algorithm.3)Propose an extension scheme of Openflow to support hierarchical software defined multi-domain optical network control technology and spectrum defragmentation technology runs on it.Apply the scheme in the testbed to accomplish function test of hierarchical software defined multi-domain optical network control technology and experiments of spectrum reconstruction to verify the feasibility of the scheme.