Algorithm For Cognitive Spectrum Optical Network Resource Optimization

With the rapid development of cloud computing,large data,and Internet of Things,IP business grows exponentially.The traditional WDM optical network is poor of flexibility,and waste of resources seriously,which can not meet the needs of large capacity,high flexibility backbone.Next-generation optical networks,spectrum flexible optical network,with high flexibility and scalability,is more rational use of spectrum resources.In the spectrum flexible optical network,control plane strategy is single,and the algorithm is fixed,which could not apply to frequent business changes issues well.However,spectrum cognitive optical network,can improve the performance of spectrum flexible optical network through cognitive style,increase its flexibility and promote their further development.The article is to study the spectrum of cognitive optical network,where to solve the spectrum resources optimization problems.This paper focuses on the cognitive spectrum optical network resource optimization problems.With the service attributes of bandwidth,intensive study of routing and spectrum allocation scheme for resource optimization.blocking traffic reducing,and network performanc improving.The main research work and results of this paper are as follows:First,considering the complex business,variabile routing,spectrum allocation policy and network efficiency of the spectrum cognitive optical networks,we propose a network architecture based on the spectrum cognitive,"Spectrum Engineering".Based on this framework,we study the routing and spectrum allocation process and corresponding implementation strategies.Second,in terms of spectrum allocation algorithm for different service bandwidth attributes,we proposed pre-planning strategy based on the spectral bandwidth of perception.In the business cognitive process,this strategy adjust spectrum allocation plan automaticly,to achieve a more rational spectrum allocation policy.Depending on the spectrum and spectrum allocation plan,we put forward Equal Spectrum Division-Section FF Spectrum Assignment(ESD-SFSA),Proportional Spectrum Division-Section FF Spectrum Assignment(PSD-SFSA),Equal Spectrum Division-Two Section Spectrum Assignment(ESD-TSA)and Proportional Spectrum Division-Two Section Spectrum Assignment(PSD-TSA)four different spectrum allocation algorithm.Simulation results show that PSD-TSA performance better.Third,in terms of route optimization algorithm,based on perceived service bandwidth and topology links,we proposed a pre-planned spectrum routing strategy.Through the way of spectrum cognitive in the routing,the spectrum feedback of link status information,timely adjustment of the routing weight-control parameters to achieve a more rational route optimization strategy.And the simulation verify the feasibility of this routing algorithm.Finally,spectrum allocation and routing of the proposed,discussed the impact of both the association and routing algorithms for optimization algorithms.Based on spectral bandwidth of property cognitive resources,we proposed a new resource optimization.And simulation experiments show that the algorithm which to improve spectrum utilization and reduce traffic blocking rate generally performance better.Also,the choice of spectrum allocation method and routing algorithm regulatory process parameters,depending on the network status and operational state.In spectrum cognitive process,we choose a different allocation method selection and regulation parameters can make the algorithm performance better,which proved that the cognitive processes is important.