This paper considers the problem of secure packet routing at the maximum achievable rate in a Quantum key distribution (QKD) network. Assume that a QKD protocol generates symmetric private keys for secure communication over each link in a multi-hop network. The quantum key generation process, which is affected by noise, is assumed to be modeled by a stochastic counting process. Packets are first encrypted with the available quantum keys for each hop and then transmitted on a point-to-point basis over the communication links. A fundamental problem that arises in this setting is to design a secure and capacity-achieving routing policy that accounts for the time-varying availability of the quantum keys for encryption and finite link capacities for transmission. In this paper, by combining the QKD protocol with the Universal Max Weight (UMW) routing policy, we design a new secure throughput-optimal routing policy, called Tandem Queue Decomposition (TQD). TQD solves the problem of secure routing efficiently for a wide class of traffic, including unicast, broadcast, and multicast. One of our main contributions in this paper is to show that the problem can be reduced to the usual generalized network flow problem on a transformed network without the key availability constraints. Simulation results show that the proposed policy incurs a substantially smaller delay as compared to the state-of-the-art routing and key management policies. The proof of throughput-optimality of the proposed policy makes use of the Lyapunov stability theory along with a careful treatment of the key-storage dynamics.
翻译:本文思考了在 Qantum 密钥分布网( QKD) 中以最大可实现速率设定安全包路径的问题。 假设 QKD 协议产生对称私密密密钥, 用于多点网络中每个链接的安全通信。 量钥生成过程受到噪音的影响, 假设该过程由随机计数程序来模拟。 包首先用每个跳上的现有量子密钥加密, 然后在通信链接的点对点基础上传输。 在此环境下出现的一个基本问题是设计一个安全和实现能力的路径处理政策, 用于计算加密和传输的有限连接能力。 在本文中, 将QKD 协议与通用 Max Weight (UMW) 路由化政策结合起来, 我们设计了一个新的安全的吞吐量最佳路径政策, 称为 Tandemem Que Decomposition( TQD) 。 QD 解决了安全路径运行问题, 并解决了一条安全路径的路径转换问题, 将连接的路径处理方法用于较宽的路径转换的路径, 将正常的路径政策用于对常规网络的流流流流压, 显示一个正常的路径管理, 将显示一个驱动到一个驱动到一个正常的路径, 将显示一个驱动到一个驱动到一个驱动到一个驱动到一个驱动到一个驱动到多点。