搜索

x
中国物理学会期刊

基于多尺度纠缠重整化假设的测量设备无关量子密钥分发协议

Measurement-device-independent quantum key distribution based on the multi-scale entanglement renormalization ansatz

PDF
导出引用
  • 测量设备无关量子密钥分发(Measurement Device Independent Quantum Key Distribution, MDI-QKD)协议因其对测量端攻击的天然免疫性而成为当前量子通信领域的研究热点,但其密钥生成依赖于两个独立光子同时干涉成功,导致成功概率随通信距离增加和信道损耗增大而急剧下降,严重限制了远距离量子通信能力。且一次成功干涉仅能提取 1 比特密钥信息,编码效率较低,使其难以满足高密钥率应用场景的需求。针对这些关键问题,本文提出将多尺度纠缠重整化假设(Multi-scale Entanglement Renormalization Ansatz,MERA)应用于测量设备无关量子密钥分发协议,并设计了三种具体实现方案:基于MERA态第L层压缩单光子态的MDI-QKD、基于MERA态第L-1层压缩纠缠态的MDI-QKD和混合模式下的MDI-QKD。这些方案利用MERA独特的分层压缩结构,每次压缩均显著减少光子数量,最终仅需传输两个甚至单个光子,即可重构出多层包含多个光子的矩阵乘积态(Matrix Product State,MPS),从而一次性实现多比特的高效密钥生成,极大地减少了量子信道的使用次数。通过动态调整压缩深度,这三种方案能够根据实际网络环境和误码率情况灵活优化量子资源使用效率,避免低层次网络资源的浪费。同时,相比经典MDI-QKD协议每次干涉仅产生1比特密钥,本文提出的方案实现了密钥规模的指数级增长。通过仿真实验,基于MERA第L层压缩单光子的MDI-QKD协议在通信距离为150km时,较经典MDI-QKD协议提高了约22931%的密钥生成率。基于MERA第L-1层压缩纠缠态的MDI-QKD和混合模式下的MDI-QKD协议,在350km处仍能保持10^-11 bit/pulse 的密钥生成率,较经典MDI-QKD协议延长了约 150 km 的有效通信距离。

    Measurement-device-independent quantum key distribution (MDI-QKD) has attracted widespread attention due to its effective defense against attacks on the detection side. However, its key generation rate is limited by the success probability of two-photon interference, which drops sharply under high channel loss and thereby limits high-throughput long-distance communication. moreover, each successful interference event yields only 1 bit of key information. To overcome these limitations, we propose the introduction of the multi-scale entanglement renormalization ansatz (MERA) into the MDI-QKD protocol. This constructs a hierarchical entanglement structure capable of efficiently encoding multi-bit information, aiming to significantly enhance the key rate, extend communication distance, and maintain theoretical security.
    We design three MERA-based MDI-QKD implementation schemes:
    (1) A single-photon scheme based on the compressed single-photon state at the L-th layer of MERA, which compresses multi-layer information into a single-photon state, enabling the extraction of multiple key bits from a single successful interference event, making it suitable for short-to-medium distance scenarios requiring high key rates;
    (2) An entanglement-based scheme based on the compressed entangled state at the (L-1)-th layer of MERA, which leverages the strong correlations of entangled states to enhance noise resistance, making it suitable for long-distance and high-loss channels;
    (3) A hybrid adaptive scheme that dynamically switches between the above two modes based on real-time channel quality, achieving optimal performance adaptation across all distances.
    All schemes employ the hierarchical compression mechanism of MERA, utilizing the disentangling operator U and the isometric mapping operator W to compress the quantum state layer by layer, and reconstructing the state via inverse operations for multi-level key extraction. The protocol incorporates the decoy-state method for eavesdropping detection, and uses basis sifting and multi-layer fidelity verification to ensure information integrity and security. Additionally, by designing four predefined MERA topological structures, efficient structural matching and information synchronization are achieved.
    Numerical simulations show that,Within the 0–200 km range, the single-photon scheme SI-MERA-MDI-QKD improves the key generation rate by approximately 22931% compared to classical MDI-QKD. The entanglement-based scheme ES-MERA-MDI-QKD and the hybrid scheme HY-MERA-MDI-QKD still maintain an effective key rate of 10^-11 bit/pulse at 350 km, extending the effective communication distance of classical MDI-QKD by approximately 150 km.
    Under varying detector efficiencies and channel loss conditions, the proposed schemes all demonstrate significant performance advantages. Coding capacity analysis shows that, under zero-error conditions, the encoding capability of MERA-MDI-QKD can exceed that of classical MDI-QKD by more than 221 times.
    Security analysis shows that the proposed schemes exhibit strong robustness against collective attacks, intercept-resend attacks, and Trojan-horse attacks. Under collective attacks, any eavesdropping behavior leads to the non-orthogonality of ancillary states, resulting in detectable errors. Under intercept-resend attacks, attackers struggle to reconstruct the original entanglement structure. Under Trojan-horse attacks, the protocol still maintains a high secure key rate through multi-layer encoding.

    目录

    返回文章
    返回
    Baidu
    map