Probing the robustness of various self-testing protocols for mulipartite entangled states

Abstract: Device-independent certification of multipartite entangled states plays a central role in a wide range of practical applications, including quantum networks, conference key agreement, and verifiable distributed quantum computation. A particularly important class of multipartite entangled states is the class of Greenberger-Horne-Zeilinger (GHZ) states. Many Bell operators have been proposed to self-test GHZ states. However, in practical scenarios, due to imperfections and the finite collection of statistics, the observed statistics do not satisfy the ideal self-testing relations. Hence, it becomes essential to investigate and compare the robustness of the different self-testing protocols. In this work, we investigate the robustness of self-testing schemes constructed from Bell operators due to Svetlichny and Mermin--Ardehali--Belinskii--Klyshko (MABK), using the analytic operator-inequality framework developed by Kaniewski [\href{https://doi.org/10.1103/PhysRevLett.117.070402}{Phys. Rev. Lett. 117, 070402 (2016)}]. We derive lower bounds on the extractable fidelity as a function of the observed value of these Bell operators. Although these protocols self-test the same underlying state, they exhibit markedly different levels of robustness. By comparing the resulting fidelity bounds, we demonstrate that the self-testing scheme based on the Svetlichny's Bell operator is the more robust among the two. Our results thus identify the Svetlichny operator based self-testing protocol as the most favorable candidate for device-independent certification of GHZ states in realistic, noisy experimental scenarios.

• The paper shows that Svetlichny-based protocols certify multipartite GHZ states with no robustness gap for four or more parties through analytic, linear fidelity bounds.
• It employs the STOPI method and operator inequalities to derive tight bounds as a function of observed Bell violations under realistic (noisy) experimental conditions.
• The study highlights that MABK-based protocols require significantly higher violation thresholds, making them less feasible than Svetlichny protocols for scalable quantum certification.

Robustness Analysis of Self-Testing Protocols for Multipartite Entangled States

Introduction

Device-independent self-testing forms the foundation for certifying multipartite entanglement in quantum networks, cryptographic protocols, and distributed quantum computation. Central to these applications are Greenberger-Horne-Zeilinger (GHZ) states, which exhibit maximal multipartite nonlocality. Certifying such states in a device-independent manner necessitates robust self-testing protocols, particularly in practical (i.e., noisy and finite-sample) experimental regimes. This paper examines the robustness of self-testing protocols based on Svetlichny and Mermin-Ardehali-Belinskii-Klyshko (MABK) Bell operators, focusing on analytic fidelity bounds as a function of observed Bell violations for up to five parties, and comparing their suitability for experimental certification of GHZ states (2605.03518).

Background and Methods

Framework and Analytical Tools

Self-testing is defined as the unique device-independent identification (up to local isometry) of a target quantum state and measurements from observed correlations. The traditional approach relies on maximal Bell inequality violation, necessitating robust extensions for realistic conditions. Previous robust self-testing methods (e.g., based on trace distance) yield loose, nonlinear bounds and are ineffective for non-ideal experimental data. Contrary to these, the operator inequality method (STOPI) by Kaniewski [Kaniewski, Phys. Rev. Lett. 117, 070402 (2016)] enables derivation of analytic, tight, linear lower bounds for the extractable fidelity as a function of Bell violation.

The authors extend Kaniewski's STOPI method to multipartite GHZ states, constructing explicit operator inequalities for generalized Svetlichny and MABK functionals. The extraction maps are defined as local dephasing channels, parameterized by the measurement settings, and calculated using only Pauli observables due to the applicability of Jordan's lemma.

Targeted Bell Scenarios

The paper analyzes multipartite Bell scenarios where each of $n$ spatially separated parties selects between two dichotomic measurements. Two classes of Bell operators are addressed:

• Svetlichny Operator: Detects genuine multipartite nonlocality; any violation implies full $n$-way nonlocality, robust against biseparable models.
• MABK Operator: General family of multipartite Bell operators, maximal violation by the GHZ state, but can exhibit large local bounds and is not sensitive solely to genuinely $n$-partite nonlocality.

For both classes, the maximal local bound ($\beta_L$) and quantum bound ($\beta_Q$) are analytically known. Self-testing proceeds by deriving a channel such that the post-processed state has high fidelity $F$ with the GHZ state as a function of observed Bell value $\beta_O$.

Robustness Results for Svetlichny-Based Protocols

Analytical Construction

For $n = 3, 4, 5$ parties, the authors construct explicit operator inequalities of the form

$$\mathcal{K}_n \succeq s\mathcal{S}_n + \mu \mathbb{I},$$

where $\mathcal{K}_n$ is the channel-adjoint-applied GHZ state, $n$0 is the parametrized Svetlichny operator, and $n$1, $n$2 are constants. By block-diagonalizing the resulting persymmetric matrices, positivity is established for all valid parameter ranges.

Tightness and Fidelity Bounds

A principal result is that, for $n$3, the analytically derived threshold Bell values at which non-trivial fidelity ($n$4) is certified coincide with the local bounds. Thus, any observed violation immediately certifies non-trivial extractability,

$n$5

with no "robustness gap" between theory and experiment for $n$6. For $n$7, a slightly nontrivial threshold persists, but the authors conjecture this can be eliminated with further channel optimization [cf. (Chen et al., 20 Mar 2026)].

Implications

This means that, for practical certification of $n$8-partite GHZ states in realistic, noisy conditions, the Svetlichny protocol is uniquely robust: any statistically significant violation suffices for device-independent GHZ certification, even as the number of parties increases.

Robustness Results for MABK-Based Protocols

Analytical Tradeoff and Thresholds

A parallel analysis for MABK functionals demonstrates that, for $n$9, the threshold Bell values required for nontrivial fidelity can substantially exceed the local bound, and this gap widens with $n$0:

• For $n$1, the threshold is $n$2 (for maximal violation by a biseparable state).
• For $n$3, the threshold is $n$4.
• For $n$5, the threshold is $n$6.

The corresponding fidelity lower bound is

$n$7

where $n$8 is the threshold above which nontrivial fidelity is guaranteed.

Comparison

Numerically, the fraction of the quantum bound needed for self-testing rapidly increases with $n$9 under MABK protocols, rendering them less viable for experimental GHZ certification at scale. For instance, for $\beta_L$0, over $\beta_L$1 of the maximal quantum violation is necessary to surpass the $\beta_L$2 fidelity threshold—a challenging requirement for state-of-the-art experiments.

Theoretical and Practical Implications

Experimental Feasibility and Quantum Networks

The demonstrated robustness of Svetlichny-based protocols underpins their practicality for robust device-independent certification of large-scale multipartite entanglement, which is crucial for quantum network security, conference key agreement, and distributed quantum computing. The MABK-based schemes, though tight in theoretical analysis, lack experimental accessibility for large $\beta_L$3 due to their disadvantageous thresholds.

Generalization and Outlook

The authors conjecture that the Svetlichny approach yields tight robust self-testing for arbitrary $\beta_L$4. The analytic techniques, notably STOPI and symmetry-based block decompositions, provide a template for robustness analyses beyond binary-outcome scenarios—for example, higher-dimensional qudits, generalized Bell inequalities, or nonprojective measurement self-testing [cf. 105.032416, (Chen et al., 20 Mar 2026)]. Future directions include extending analytic robustness to the elegant Bell inequality, chained inequalities, and prepare-and-measure settings, with direct implications for device-independent random number generation and cryptographic protocols.

Conclusion

By systematically analyzing and comparing Svetlichny and MABK-based self-testing protocols, the paper establishes that Svetlichny inequalities uniquely enable robust, device-independent certification of multipartite GHZ states in realistic experimental settings. This result is both theoretically tight and experimentally relevant for scalable quantum networking and cryptography. The analytic approach introduced provides a foundation for further extensions to more general scenarios in device-independent quantum information science (2605.03518).