Nonlocality in Continuous-Variable Quantum Networks

TL;DR

This paper develops a pseudospin-based framework to study nonlocality in continuous-variable quantum networks, focusing on linear-chain and star topologies. It derives explicit maximal violations of nonlinear $n$-local inequalities for arbitrary two-mode CV states and analyzes Gaussian (TMSV) as well as non-Gaussian resources, including photon-subtracted states, entangled coherent states, and CV Werner states. A key result is that star networks with identical TMSV resources exhibit violations that are independent of network size, while linear chains see diminishing violations with more links; non-Gaussian resources can substantially enhance nonlocal correlations and even enable maximal violation at zero squeezing in certain configurations. The work also proposes an experimentally viable route using spatial parity observables, bridging theory and experiment in CV network nonlocality and suggesting directions for scalable quantum networks and quantum information protocols.

Abstract

Quantum networks enable forms of nonlocality beyond the standard Bell scenario, with a multitude of potential applications. Continuous-variable (CV) platforms are particularly attractive for large-scale networks, offering deterministic entanglement generation and favorable prospects for long-distance distribution. Here we present a formalism to study CV network nonlocality using pseudospin measurements. Considering the linear chain and star configurations, we derive the maximal violations of the corresponding network locality inequalities for arbitrary two-mode states. Using two-mode squeezed vacuum states, we show that the strength of nonlocality in the star configuration remains independent of the network size. Moreover, the nonlocal correlations persist even at arbitrarily high temperatures provided the squeezing exceeds a critical threshold. Further, we demonstrate non-Gaussianity as an enhancer of network nonlocality through illustrations of various classes of non-Gaussian resources. Remarkably, a coherent superposition of single-photon subtractions across modes achieves maximal violation for vanishing squeezing. Finally, we provide schematics of an experimentally feasible implementation of CV network nonlocality based on the isomorphism between pseudospin and spatial parity observables.

Figures (10)

• Figure 1: Bilocal network scenario
• Figure 2: Linear Chain network scenario
• Figure 3: Star network scenario
• Figure 4: Maximal $n$-locality expressions for linear-chain and star network configurations using TMSV states as a function of the squeezing parameter $K = \tanh(2r)$, assuming equal squeezing $r_i = r$ for all sources. The violations increase monotonically with squeezing and approach the maximal quantum value $2\sqrt{2}$ in the limit $K \rightarrow 1$. The horizontal dashed lines at $S = 2$ (green) and $S = 2\sqrt{2}$ (gray) denote the classical and maximal quantum bounds, respectively.
• Figure 5: $S^{\max}_{\rho_{*}, \rho_{*}}$ is plotted as a function of $p$ and $K$, for fixed inverse temperature parameters $\beta_1 = \beta_2 = 1$. The red contour line corresponds to $S^{\max}_{\rho_{*}, \rho_{*}} = 2$, separating the bilocal ($S^{\max}_{\rho_{*}, \rho_{*}} \leq 2$) and nonbilocal ($S^{\max}_{\rho_{*}, \rho_{*}} > 2$) regions in the parameter space.

Theorems & Definitions (4)

• Theorem 1
• proof
• Theorem 2
• proof