[Paper Review] Bell inequalities: many questions, a few answers

Nicolas Gisin|arXiv (Cornell University)|Feb 2, 2007

Philosophy and History of Science4 citations

TL;DR

This paper presents a new family of Bell inequalities for bipartite systems with arbitrary even numbers of settings and binary outcomes, demonstrating that quantum mechanics can violate these inequalities by a factor of √m using maximally entangled states and optimized measurements. The key contribution is a novel inequality—S₃ₓ₄—that achieves a quantum violation of 4√3 ≈ 6.928 for n=3, m=2, exceeding classical limits and revealing nonlocal correlations requiring complex Hilbert spaces.

ABSTRACT

What can be more fascinating than {\it experimental metaphysics}, to quote one of Abner Shimony's enlightening expressions? Bell inequalities are at the heart of the study of nonlocality. I present a list of open questions, organised in three categories: fundamental; linked to experiments; and exploring nonlocality as a resource. New families of inequalities for binary outcomes are presented.

Motivation & Objective

• To address open questions in quantum nonlocality by proposing new families of Bell inequalities for binary outcomes and arbitrary even numbers of settings.
• To explore the role of nonlocality as a quantum resource, particularly in quantum information processing and communication.
• To investigate the geometry of local polytopes and the mismatch between classical convex hulls and quantum symmetries in high-dimensional systems.
• To demonstrate that quantum strategies can surpass classical limits by a factor of √m using maximally entangled states and projective measurements in mutually conjugated bases.

Proposed method

• Derives a new class of Bell inequalities using conditional probabilities under the locality assumption, defined via coefficients C_{abc...}^{xyz...} and a local bound S_lhv.
• Proposes a quantum strategy where Alice and Bob share a maximally entangled state of dimension m, with Alice measuring in two mutually conjugated bases depending on her input x.
• Employs Bob’s measurement as a projection onto an intermediate state between two bases, with outcome b=1 indicating success and b=0 failure, thus defining a score function.
• Computes the quantum score S_quantum = 2√m for n=2, and S_quantum = 4√3 ≈ 6.928 for n=3, m=2, showing violation of classical bounds.
• Represents the inequality S₃ₓ₄ using a 4×3 matrix notation and visualizes optimal measurement settings on the Poincaré sphere, with Alice’s settings as orthogonal vectors and Bob’s as tetrahedral vertices.
• Analyzes the geometry of the local polytope and shows that the new inequality S₃ₓ₄ is not a facet, indicating a mismatch between local polytope structure and quantum symmetry.

Experimental results

Research questions

• RQ1 Can new families of Bell inequalities be constructed for arbitrary even numbers of settings and binary outcomes that reveal stronger quantum violations?
• RQ2 To what extent can quantum nonlocality be harnessed as a resource in quantum information protocols, particularly in communication and computation?
• RQ3 Does the use of complex Hilbert spaces enable higher quantum violations than those achievable with real numbers in higher-dimensional systems?
• RQ4 Why do certain elegant quantum inequalities, such as S₃ₓ₄, fail to be facets of the local polytope despite their symmetry and high quantum violation?

Key findings

• For n=2 and m=2, the new inequality reduces to the CHSH inequality, with a quantum violation of 2√2 ≈ 2.828, confirming consistency with known results.
• For n=2 and general m, the quantum score reaches S_quantum = 2√m, exceeding the classical local bound S_local = 2 by a factor of √m.
• For n=3 and m=2, the quantum maximum is S_quantum = 4√3 ≈ 6.928, significantly exceeding the classical bound of 6.
• Numerical evidence suggests that the maximum quantum violation using only real numbers is lower—e.g., 10/3 ≈ 3.333 for m=3 and 2+2√5 ≈ 6.472 for n=3, m=2—indicating that complex Hilbert spaces may be essential for optimal nonlocality.
• The inequality S₃ₓ₄, represented as a 4×3 matrix, is not a facet of the local polytope, revealing a geometric mismatch between classical convex hulls and quantum measurement symmetries.
• Optimal measurement settings for S₃ₓ₄ are visualized on the Poincaré sphere: Alice’s three settings are mutually orthogonal, and Bob’s four are located at the vertices of a tetrahedron.