Modeling And Optimization Of Quantum Key Distribution Network

Quantum Key Distribution(QKD)technology can provide secure key distribution service for remote communication parties,based on the fundamental principles of quantum mechanics.Combining this technology with encryption algorithms,such as one time pad,can greatly improve the security of core message transmission in cyberspace.However,the specificities of QKD system,for example,it can only support key distribution between two nodes,and the key distribution rate is greatly limited by the length of optical fiber,etc.,make it difficult to construct a high-performance QKD network with multiple QKD systems.In this dissertation,aiming to maximize the satisfaction degree of communication demand,the high-performance QKD network construction is studied,according to the description and simulation of critical features,from the two perspectives of static topology design and dynamic routing optimization.This research can evidently promote the largescale application of QKD technology in the field of cyberspace security,both theoretically and practically.The main works and contributions of this dissertation are as follows.(1)A SHQN mathematical model based on network flow theory is established,to describe the five static critical features of hybrid QKD networks mathematically,namely,the variety of QKD protocols,the limitation of key distribution rate,the necessity of untrusted-relay selection,the contradiction between key supply and demand and the concurrency of communication demand,which lays a theoretical foundation for the follow-up research.Specifically: according to the design of protocol classification method based on fiber dependency,a topological representation including edge classification and flow classification is designed to describe the variety of QKD protocols mathematically;An edge attribute named key distribution rate is designed to describe the limitation of key distribution rate mathematically;Three node attributes named average communication demand,key consumption ratio,and credibility are designed to describe the concurrency of communication demand mathematically;Three flow attributes named integer-multiple limitation,capacity limitation and flow conservation are designed to describe the contradiction between key supply and demand mathematically.(2)A static topology evaluation method and the corresponding optimal hybrid topology design scheme are proposed,based on the SHQN mathematical model proposed above,to improve the average satisfaction degree of communication demand from the perspective of topology.In terms of topology evaluation: a topology evaluation method including ITS communication capability and network construction cost is proposed to balance the game relationship between the complex communication demand and the credibility control cost;Based on two typical topology construction tasks of QKD system placement and intermediate node selection,the effectiveness and necessity of the proposed topology evaluation method are fully verified.In terms of topology design: a hybrid topology design scheme is proposed,by setting the maximal ITS communication capability as optimization objective,cost limitation,credibility control,etc.,as constraint conditions;Based on a real NSFNET topology and lots of virtual random topologies,the effectiveness and universality of the proposed topology design scheme are fully verified.(3)A DHQN simulation model based on hybrid topology is proposed,on the basis of proposed SHQN mathematical model,to further simulate the dynamic critical feature,namely,the volatility of communication demand.Specifically: by choosing the application mode of data-relay and designing the two-layer architecture with data-key separation,a simulation framework with routing protocol as the core is proposed to describe the contradiction between key supply and demand accurately;A key distribution module based on logical topology transformation is designed to describe the variety of QKD protocols,the limitation of key distribution rate,and the necessity of untrusted-relay selection accurately;A traffic generation module based on Poisson stochastic process is designed to describe the volatility of communication demand and the concurrency of communication demand accurately.(4)A dynamic routing evaluation method and two routing optimization schemes are proposed,based on the DHQN simulation model proposed above,to improve the real-time satisfaction degree of communication demand.In terms of routing evaluation: a routing evaluation method including ITS communication efficiency,packet delivery rate,oneway delay and routing cost is proposed,by taking the ITS communication capability as a comparison and the communication demand satisfaction as a benchmark;Considering the similarity between wireless ad-hoc networks and QKD networks,the effectiveness of the proposed routing evaluation method is verified by the comparison of three typical wireless ad-hoc network routing protocols.In terms of routing optimization: two optimization strategies of the link-state prediction based on remaining key quantity and the pathfinding strategy improvement based on key recovery capability are proposed,to solve the problems of frequent change with link states and low utilization with quantum keys respectively;The advantages of the proposed optimization strategies is demonstrated by the comparison of improved routing protocol and two existing routing protocols.