Quantum key distribution (QKD) technology provides proved unconditional security in pointto-point key transmissions based on quantum mechanics. However, several limitations, most prominently range, make it complex to construct large QKD network. The trusted relay QKD network is the most practical scene, and stochastic routing scheme was proposed to solve the fatal problem that some relay nodes may be controlled by adversary. But most results in stochastic routing scheme hinge on an adversary model without consideration of attacker's prior-knowledge and attack strategy. To reveal the security problem, we propose a new adversary model and apply it in grid-shaped partially-trusted relay QKD network. Then we analyze the best attack strategy and the impact on security under various situations respectively. Analysis and simulation results show that in some cases the security of stochastic routing scheme will be compromised, and a generalized upper bound for the overall safe probability is obtained. We also discuss how to alleviate the security problem, making the network more robust to attacks.
Quantum key distribution (QKD) technology provides proved unconditional security in pointto-point key transmissions based on quantum mechanics. However, several limitations, most prominently range, make it complex to construct large QKD network. The trusted relay QKD network is the most practical scene, and stochastic routing scheme was proposed to solve the fatal problem that some relay nodes may be controlled by adversary. But most results in stochastic routing scheme hinge on an adversary model without consideration of attacker's prior-knowledge and attack strategy. To reveal the security problem, we propose a new adversary model and apply it in grid-shaped partially-trusted relay QKD network. Then we analyze the best attack strategy and the impact on security under various situations respectively. Analysis and simulation results show that in some cases the security of stochastic routing scheme will be compromised, and a generalized upper bound for the overall safe probability is obtained. We also discuss how to alleviate the security problem, making the network more robust to attacks.
In process of deploying large-scale quantum key distribution(QKD) networks, in order to solve the problems of service congestion, key waste and uneven resource allocation in key relay, inspired by the behaviour of fish-like intelligent clusters in nature, this paper proposes a simulated fish stochastic routing scheme suitable for partially-trusted relay QKD network scenarios. Firstly, the routing process of key distribution is regarded as the optimization behavior of fish swarm, and the simulated behavior model of fish swarm under the key transmission mechanism is constructed. Secondly, the distance factor and routing cost are incorporated to the food concentration of the node as a routing evaluation index, this paper establishes the ILP routing model based on the maximum food concentration and the fish-sense routing model. Finally, in order to deploy the network resources, a stochastic routing algorithm based on perceived fish swarm for QKD network is designed. Simulation results show that this scheme can effectively improve the quantum key distribution success rate, and the average key consumption of the optical path is less. When attackers adopt the equal probability attack strategy, the overall security probability of the scheme is improved by 8% and 2% compared to the Random and RIP-SRA algorithms. In terms of link load uniformity, the labeled diffusion behavior can effectively balance the key on QKD network. The standard deviation of the average key residual is 270.2673, and the key consumption tends to be levelled off.
The long-distance transmission of quantum secret key is a challenge for quantum communication. As far as the current relay technology is concerned, the trusted relay technology is a more practical scheme. However, the trusted relay technology requires every relay node to be trusted, but in practical applications, the security of some relay nodes cannot be guaranteed. How to overcome the security problem of trusted relay technology and realize the security key distribution of remote quantum network has become a new problem. Therefore, in this paper, a method of quantum key distribution in ring network is proposed under the condition of the coexistence of trusted and untrusted repeaters, and proposes a partially-trusted based routing algorithm (PT-RA). This scheme effectively solves the security problem of key distribution in ring backbone network. And simulation results show that PT-RA can significantly improve key distribution success rate compared with the original trusted relay technology.
Quantum key distribution (QKD) can protect the exchange process of confidential information between communicating parties. By using the basic principles of quantum mechanics and combined with “one-time pad” cipher encryption, information can be unconditionally secure. The BB84 protocol first describes the method of transmitting information by photon polarization state, and it expounds the transmission process of services between trusted relays. However, due to the defects of real experimental devices, there are security vulnerabilities in QKD in a real system. The birth of measurement-device-independent quantum key distribution (MDI-QKD) protocol solves the problem, providing immunity to hacker attacks at the end of the detector. It can enable both sides of the transmission service to establish a connection and generate secret keys through an untrusted relay node to ensure information security. However, the types and properties of link nodes in quantum key distribution network (QKDN) based on partially trusted relay are more complex, which can easily result in network fault. Therefore, how to prevent the impact of failure on QKDN has become an urgent problem. In this paper, we propose a protection scheme for QKDN with partially trusted relays. The method deals with trusted and untrusted relays differently and constructs the working and protection paths of the secret key for each service. It reduces resource conflict between the protection and working paths by establishing a key protection threshold, which realizes the resource trade-off between the two factors. Simulation results show that the scheme provides effective protection to the services, and it improves the stability and reliability of QKDN based on partially trusted relay.
ABSTRACT Due to distance and key‐rate limitations, current quantum key distribution (QKD) technology requires the use of trusted nodes, which, together with QKD links, form a graph referred to as a QKD network. We address the problem of securely routing (relaying) keys between trusted nodes, where some of them can be malicious. Existing solutions typically rely on centralized controllers, which simplify adaptive routing and congestion control but introduce single points of failure. Moreover, some proactive key‐relay protocols distribute keys to multiple nodes in the network; thus, compromising a single node results in the compromise of the shared keys. In this paper, we propose a decentralized, topology‐oblivious key‐relay algorithm based on random walks. Our algorithm provides information‐theoretic security (ITS) for QKD network topologies having at least one non‐malicious path between any pair of nodes (in particular, ‐vertex‐disjoint paths can tolerate malicious cartels of size ). Besides, it provides strong computational security guarantees for other topologies (even for graphs with malicious articulation points). We evaluate the proposed algorithm on the GÉANT GN4‐3N topology, which provides a realistic approximation of a multi‐state cross‐border topology for future QKD networks.
Quantum Key Distribution (QKD) provides unconditional peer-to-peer security based on the principles of quantum physics. By utilizing relay nodes, the security of QKD can be extended over longer distances. However, the introduction of relay nodes brings both security and communication success rates issues. To tackle those issues we propose an enhanced multi-path scheme. The key features of our proposal are as follows: 1. By taking the reliability of relay nodes as one of the algorithm inputs,making the scheme more suitable for partially trusted QKD (PTQKD) networks. 2. By using Multi-Segment Multi-Path approach increases the difficulty for attackers to obtain complete key information and improves the security of PTQKD. 3. The adaptive routing algorithm generates a sufficient number of diverse paths based on node contribution rate, key freshness, and reliability. We conducted a theoretical analysis of the algorithm,and simulation results on PTQKD demonstrate that our method outperforms traditional QKD methods in terms of security and transmission success rate. This advancement has the potential to enhance the adoption of QKD networks.
There is a need to ensure unconditionally secure communications between different sites of an organization. We use quantum key distribution (QKD) protocol to provide a continuous stream of secret key bits to each communicating pair for message encryption through one-time-pad (OTP) cryptography. Due to long distance between different sites, trusted nodes are used to relay each service key in a hop-by-hop manner from one end node to another end node of a communicating pair. In a QKD network, routes for key relaying can be affected by availability of relay key bits at trusted nodes. Also, relay key allocation among different routes can impact security. We consider the scenario of supporting multiple communicating pairs, where each pair demands delivery of a certain service key rate across the QKD network, and trusted nodes may not have sufficient relay key rates to perform bitwise encryption to protect the service key. Hence, there is a need to allocate the limited relay key bits to meet all service key delivery demands while considering its effects on security. We propose to perform jointly routing and relay key allocation to minimize attack success probability. We have formulated the joint problem as an optimization and further proposed an efficient algorithm to find the optimal solutions. We have evaluated the proposed algorithm to confirm its ability and efficiency in finding optimal routes and relay key allocations. For the specific network topology that we have adopted, results show that the minimum number of required trusted nodes increases linearly when end nodes are added to the system. Furthermore, we can achieve a higher security level by adding more trusted nodes in a QKD network. But increasing the number of trusted nodes beyond the minimum requirement may not continue to reduce significantly the attack success probability.