Bilocal Network Experiment Overview

• Bilocal network experiments are defined by two independent sources connecting three nodes, revealing non-classical correlations through entanglement swapping.
• The experimental methodology employs polarization-entangled photons and optimized measurement settings to certify violations of bilocality inequalities.
• Classical bilayer network studies use analytic models to quantify interlayer topological differences, linking shared node fractions with network robustness.

A bilocal network experiment investigates non-classical correlations in quantum and classical networks featuring two independent sources and a specific node-sharing or entanglement-swapping geometry. In quantum contexts, these experiments reveal nonbilocal correlations violating constraints imposed by local realism augmented with source independence, as formalized by bilocal hidden variable models. In classical multilayer networks, the "bilocal" paradigm typically addresses systems with two interdependent layers sharing a subset of nodes (“interconnecting nodes”) and analyzes how the number and positions of these nodes relate to structural and functional differences between layers. This article reviews the foundational definitions, analytic models, empirical methodologies, and principal findings characterizing bilocal network experiments in both quantum and classical settings.

1. Formal Definition and Conceptual Framework

Bilocal network experiments, originally proposed to extend Bell-type nonlocality studies to networked quantum systems, model networks as consisting of three nodes (parties)–commonly labeled Alice (A), Bob (B), and Charlie (C)–connected by two independent sources. In the quantum scenario, S₁ produces entanglement between A and B, and S₂ between B and C. The critical assumption of source independence leads to bilocal hidden variable models specified by two independent variables λ₁ and λ₂. This yields the decomposition

$$p(a,b,c|x,y,z) = \int dλ_1 dλ_2~ ρ_1(λ_1) ρ_2(λ_2)~ p_A(a|x,λ_1)~ p_B(b|y,λ_1,λ_2)~ p_C(c|z,λ_2)$$

where x, y, z denote measurement choices and a, b, c are outcomes at parties A, B, C respectively (Andreoli et al., 2017).

In classical multilayer network settings, the "bilocal" or "bilayer" network is constructed by two graphs G¹ and G², with nonempty overlap V₁ ∩ V₂ of "interconnecting nodes" that simultaneously appear in both layers, serving potentially different roles in each (Qu et al., 2010).

2. Analytical Results and Bilocality Inequality

Quantum Bilocality Inequality

Branciard et al. derived a nonlinear “bilocality” inequality for the scenario above:

$\mathcal{B} = \sqrt{|I|} + \sqrt{|J|} \leq 1$

with

$$I = \frac{1}{4} \sum_{x,z=0}^{1} \langle A_x B_0 C_z \rangle,\quad J = \frac{1}{4} \sum_{x,z=0}^{1} (-1)^{x+z} \langle A_x B_1 C_z \rangle$$

$$\langle A_x B_y C_z \rangle = \sum_{a,b,c} (-1)^{a+b+c} p(a,b,c|x,y,z)$$

Violations of this bound experimentally certify nonbilocal correlations inaccessible to standard local hidden variable models even when only assuming independent sources (Branciard et al., 2011, Andreoli et al., 2017, Saunders et al., 2016).

Classical Bilayer Analytic Model

In interconnecting bilayer network theory, key parameters include the averaged normalized topological difference of interlayer-shared nodes for property x (degree, betweenness, etc.):

$$U_x = \frac{1}{m} \sum_{(i,j)} \left| \frac{x^1_i}{\langle x^1 \rangle} - \frac{x^2_j}{\langle x^2 \rangle} \right|$$

and the normalized count of interconnecting nodes

$n = \frac{m}{M_1 + M_2 - m}$

where m = |V₁ ∩ V₂|, and M₁, M₂ are layer sizes. Analytic models predict a monotonic dependence: U_x increases and n decreases as interlayer “function difference” grows (Qu et al., 2010).

3. Experimental Realizations and Methodologies

Quantum Bilocal Network Experiments

State Preparation and Setup.

Experiments utilize two independent polarization-entangled photon-pair sources, typically based on Type-II SPDC, simultaneous or spatially separated pumping, and strict isolation to prevent shared phase memory between sources (Andreoli et al., 2017, Saunders et al., 2016).

Measurement Schemes.

Measurement choices at A and C use settings such as A₀,C₀=(σ_x+σ_z)/√2, A₁,C₁=(σ_x–σ_z)/√2, implemented by HWP+PBS modules. Bob may be decomposed into two single-qubit stations B^A and B^C, each performing Pauli observables (e.g., σ_x⊗σ_x or σ_z⊗σ_z), with outcomes combined by XOR (Andreoli et al., 2017).

Detection and Data Analysis.

Fourfold coincidences are collected across all stations. Correlators are derived from joint probabilities; bilocality parameters are computed per inequality definitions. Error estimates typically assume Poissonian counting statistics (Andreoli et al., 2017).

Results.

Typical experiments report violation of the bilocal bound by multiple standard deviations, e.g., ℬ_exp = 1.2712 ± 0.0042 (65σ above the bound) (Andreoli et al., 2017), with observed noise levels yielding visibilities V ≈ 0.90.

Classical Bilayer Network Experiments

Model Construction and Simulation.

Two layers are generated with shifted power-law degree distributions; interconnecting nodes are assigned probabilistically or deterministically as a function of functional mismatch θ. Analytic formulas for Uₓ vs. n are derived and tested numerically (Qu et al., 2010).

Empirical Studies.

Eight real-world bilayer networks were analyzed (herb–food, biology–physics keywords, actors, yeast and E.coli PPI–metabolic, Chinese transportation). For each, normalized difference statistics and interconnectivity were computed and shown to concord with model predictions across all cases (Qu et al., 2010, Qu et al., 2010).

4. Interpretation, Robustness, and Limitations

Device Independence and Reference Frames

Early quantum bilocality tests often required shared reference frames for polarization, which could permit spurious “false positive” violations if alignment channels introduced hidden correlations. The experiment of Carvacho et al. (Andreoli et al., 2017) implemented reference-frame-free schemes, demonstrating robust violations even with arbitrary misalignments and showing that neither A–C nor any A/B/C pair need share a common frame.

Noise, Decoherence, and Mixed States

Noise from amplitude damping and decoherence suppresses nonbilocal correlations. Nevertheless, weak measurements and post-selection can significantly delay the decay of nonbilocality, restoring violations for higher noise strengths than otherwise possible (Gupta et al., 2018). Remarkably, certain combinations of entangled and separable mixed Werner states can also yield nonbilocality, so long as the product of their visibilities exceeds threshold v₁v₂ ≥ 0.2422 (Yang et al., 2022). This regime includes cases where one source is fully separable.

Scaling and Real-World Networks

Quantum network configurations have been extended to multi-source, multi-output structures (e.g., star, chain topologies). Star or N-locality scenarios generalize the bilocal scenario and exhibit violations with suitably adapted nonlinear inequalities (Tavakoli et al., 2014). The classical (bilayer) paradigm generalizes to arbitrary numbers of overlapping nodes and accommodates empirical systems as diverse as biological, infrastructural, or social networks. The Uₓ vs. n scaling holds universally across empirical domains (Qu et al., 2010, Qu et al., 2010).

5. Implications and Applications

Quantum Information

Bilocal network experiments establish the possibility of device-independent certification of source independence and network nonlocality, fundamental for the secure deployment of future quantum communication architectures, including repeater chains and distributed quantum computation. Removing the need for trusted reference frames addresses a scaling barrier for large networks (Andreoli et al., 2017).

Network Theory and Complex Systems

In classical domains, bilayer network analysis validates interdependent-layers as a quantitative framework for multi-functional systems. The scaling predictions for topological difference and shared-node fraction provide practical proxies for functional distance and robustness design in infrastructure, multiplex social or biological networks (Qu et al., 2010).

Future Directions

Quantum implementations are pushing toward urban-scale, multi-technology demonstrations, integration with quantum memories, and provisions for fair-sampling/detection-loophole closure. Classical bilayer work suggests investigation of temporal evolution, adaptive interconnections, and feedback between function mismatch and structural adaptation (Qu et al., 2010).

6. Summary Table: Quantum Bilocal Experiment Critical Features

Feature	Quantum Bilocal Experiment (Andreoli et al., 2017)	Classical Bilayer Networks (Qu et al., 2010)
Source configuration	Two independent EPR (SPDC) sources	Two network layers, some shared nodes
Key assumption	Independence of λ₁, λ₂ (sources)	Assignment/merging of interconnecting nodes
Measured quantity	Bilocality parameter ℬ (nonlinear)	Uₓ: normalized average topological difference
Bound for classical/bilocal model	ℬ ≤ 1 (nonlinear inequality)	Uₓ–n monotonic scaling
Typical quantum/empirical outcome	ℬ_exp ≈ 1.27 (clearly violates bound)	Empirical Uₓ vs. n matches analytic model
Influence of noise	Reduces visibility, can be countered by weak measurement	Increases topological difference, decreases n
Reference frame / alignment	Not required in RF-free scheme	Not applicable
Impact	Device-indep. network nonlocality, quantum certification	Functional interdependence, network robustness

This comparative summary highlights the core elements and distinctions between quantum and classical bilocal network experiments.

• Carvacho et al., "Experimental bilocality violation without shared reference frames" (Andreoli et al., 2017)
• Qu et al., "Study on some interconnecting bilayer networks" (Qu et al., 2010)
• Qu et al., "Empirical study on some interconnecting bilayer networks" (Qu et al., 2010)
• Saunders et al., "Experimental demonstration of non-bilocal quantum correlations" (Saunders et al., 2016)
• Branciard et al., "Bilocal versus non-bilocal correlations in entanglement swapping experiments" (Branciard et al., 2011)
• Choudhury et al., "Preservation of quantum non-bilocal correlations in noisy entanglement-swapping experiments using weak measurements" (Gupta et al., 2018)
• Wen et al., "Violating the bilocal inequality with separable mixed states in the entanglement-swapping network" (Yang et al., 2022)