Experimental Efficient Source-Independent Quantum Secret Sharing against Coherent Attacks

TL;DR

This work tackles secure multiparty quantum secret sharing without multipartite entanglement by implementing a resource-efficient SI QSS protocol that uses entangled photon pairs and a postmatching technique to emulate GHZ correlations. An experimental demonstration with two Sagnac-type sources in a star topology achieves high key rates for 3–5 users, with rates largely independent of the number of participants under equal losses. The results are analyzed under composable security with finite-key considerations against coherent and participant attacks, and they illustrate scalable, high-rate QSS suitable for large quantum networks. The approach reduces hardware complexity and offers a practical pathway toward widespread deployment of SI QSS in multiuser quantum networks.

Abstract

Source-independent quantum secret sharing (SI QSS), while essential for secure multiuser cryptographic operations in quantum networks, faces significant implementation challenges stemming from the inherent complexity of generating and distributing multipartite entangled states. Recently, a resource-efficient SI QSS protocol utilizing entangled photon pairs combined with a postmatching method has been proposed to address this limitation. In this Letter, we report an experimental demonstration of this protocol using high-fidelity polarization-entangled photon pairs in a star topology. For a three-user network, we obtain secure key rates of 21.18, 4.69, and 1.71 kbps under single-user channel losses of 7.6, 10.9, and 12.9 dB respectively. Furthermore, under conditions of equal channel loss per user, we achieve secure key rates of 6.97, 6.46, and 5.88 kbps for three-, four-, and five-user scenarios respectively. These results demonstrate the advantageous independence of the key rate from the number of users. Our work paves the way for large-scale deployment of SI QSS in multiuser quantum networks.

Figures (7)

• Figure 1: Scheme of the resource-efficient SI QSS protocol. In the physical connection layer, the untrusted central node distributes the entangled photon pairs between the dealer $A$ and the player $B_1$, and between the dealer $A$ and the player $B_2$, respectively. In the quantum correlation layer, the dealer $A$ shares the Bell states $\ket{\psi^{-}}$ with each player $B_j$ ($j=1,2$). With the conduction of the postmatching method, all participants share the equivalent GHZ-state $\ket{\Phi^{+}_{3}} = (\ket{000} + \ket{111}) / \sqrt{2}$ correlations among their classical measurement results.
• Figure 2: Experimental setup. The untrusted central node is equipped with two Sagnac-type entanglement sources to distribute Bell states between the dealer $A$ and the player $B_j$ ($j=1,2$). All protocol participants are equipped with polarization measurement modules to analyze the polarization of received photons. DPBS, dichroic polarization beam splitter; QWP, quarter-wave plate; DHWP, dichroic half-wave plate; DM, dichroic mirror.
• Figure 3: Experimental results of the resource-efficient SI QSS. We plot the experimental data in triangle scatters for $p_x = 0.9$ and square scatters for $p_x = 0.5$. Channel loss represents the average loss between the untrusted central node (entanglement sources) and the participants (detection channels of SNSPD).
• Figure 4: Schematic for the generation of GHZ state through quantum operations in three-participant scenario, which is equivalent to our experimental demonstration. The Bell state $\ket{\psi^-}$ is transformed into $\ket{\phi^+}$ through the local unitary operations $Z$ gate and $X$ gate. Then CNOT gates are performed on two Bell states $\ket{\phi^{+}}_{a_1b_1} \otimes \ket{\phi^{+}}_{a_2b_2}$ and the three-party GHZ state $\ket{\Phi^+_3}$ is established among $a_1$, $b_1$ and $b_2$.
• Figure 5: Experimental results of polarization-entanglement visibility with different average numbers of photon pairs. The interference fringes are represented in purple for the rectilinear basis and in orange for the diagonal basis. Subfigure (a) and (b) present the interference fringes of source 1 with $\mu = 0.004$ and $\mu = 0.023$, respectively. Subfigure (c) and (d) present the entanglement visibility of source 2 with $\mu = 0.002$ and $\mu = 0.021$, respectively. The polarization angle is twice the orientation angle of the fast axis of HWPs.