Research On Resource Allocation And Optimization Of Key-driven Optial Networks

In recent years,threats to security in optical networks attract more and more attention with more and more confidential and sensitive information in critical areas(government,military and finance etc.)being transferred in optical networks.As a theoretically absolutely secure key distribution scheme,quantum key distribution(QKD)technology is rapidly developing and gradually moving toward industrialization.Introducing QKD into optical networks can greatly improve the security of optical networks.Key-driven optical networks(KDON)refers to a new type of optical networks that provides secure keys for physical layer based on QKD technology.During the process of KDON,the introduction of QKD increases the dimension of network resources and makes network architecture and resource scheduling more complicated.How to perceive the services security requirements from the whole network perspective and realize the efficient scheduling of multi-dimensional resources for keys and services adaptation is the main challenge for KDON.Based on the demand of anti-interception of high-speed information in KDON,this subject overcomes the optimal scheduling of optical networks resources for key and service adaptation from the aspects of network architecture,security policy and resource allocation.Major accomplishments and innovations are as follows:First of all,aiming at the problem of centralized management and control of multi-layer multi-dimensional resources in KDON,an openFlow-enabled software-defined key-driven optical network(SD-KDON)architecture is designed to realize the programmable control of multi-dimensional network resources and security policies.The control interface and protocol are designed,and security services control flow under the architecture is detailed descripted.Secondly,focusing on the demand of cross-layer mapping between multi-dimensional resources and multi-level security services in KDON,a security strategy based on key update is designed and a multi-dimensional RWTA algorithm based on multiple security policies is proposed.Simulation results show the proposed RWTA algorithm can get better key quantification index for higher-level security services,e.g.,in the case of simulation,the key success rate of services with security level 5 is increased by 6.28%compared with services with security level 4,and the key update delay reduced by 0.11 timeslots,enabling effective adaptation of multi-dimensional network resources to services security requirements in KDON.Finally,in order to reduce fragmentation of wavelength-timeslot resources in RWTA process,the conception of time continuous compactness is proposed and three resource planning strategies are formulated.Simulation results show that RWTA schemes based on time continuous compactness can achieve lowest blocking probability,which can be reduced by 7.62%and 2.81%respectively compared with random-fit and first-fit strategy when traffic load equals 90.