Uniqueness of quantum states compatible with given measurement results

Jianxin Chen, Hillary Dawkins, Zhengfeng Ji, Nathaniel Johnston, David W. Kribs, Frederic W. Shultz, Bei ZengPhysical 2013 · DOI · arXiv

Abstract

We discuss the uniqueness of quantum states compatible with given measurement results for a set of observables. For a given pure state, we consider two different types of uniqueness: (1) no other pure state is compatible with the same measurement results and (2) no other state, pure or mixed, is compatible with the same measurement results. For case (1), it was known that for a $d$-dimensional Hilbert space, there exists a set of $4d-5$ observables that uniquely determines any pure state. We show that for case (2), $5d-7$ observables suffice to uniquely determine any pure state. Thus, there is a gap between the results for (1) and (2), and we give some examples to illustrate this. Unique determination of a pure state by its reduced density matrices (RDMs), a special case of determination by observables, is also discussed. We improve the best-known bound on local dimensions in which almost all pure states are uniquely determined by their RDMs for case (2). We further discuss circumstances where (1) can imply (2). We use convexity of the numerical range of operators to show that when only two observables are measured, (1) always implies (2). More generally, if there is a compact group of symmetries of the state space which has the span of the observables measured as the set of fixed points, then (1) implies (2). We analyze the possible dimensions for the span of such observables. Our results extend naturally to the case of low-rank quantum states.

Introduction

In quantum mechanics, fully describing a state $\rho$ in a $d$-dimensional Hilbert space typically requires $d^2 - 1$ independent measurements—a process known as quantum tomography. However, if we have prior knowledge that a state is "pure" (represented by a single vector rather than a statistical mixture), we can get away with far fewer measurements.

The authors distinguish between two types of uniqueness:

  1. UDP (Uniquely Determined among Pure states): The measurement results are compatible with only one pure state, but there might be a "mixed" state (a statistical blend) that yields the same results.
  2. UDA (Uniquely Determined among All states): The measurement results are compatible with exactly one state, period. This is much stronger because it doesn't require the researcher to assume the state is pure; the data itself proves it.

The central challenge addressed is: How many measurements are actually needed for UDA, and how does this compare to UDP?

Methods and Theoretical Framework

The authors employ a geometric approach to quantum states, treating the set of all density matrices as a convex set in a $(d^2-1)$-dimensional real space. Key techniques include the analysis of the Joint Numerical Range, which treats the set of possible measurement outcomes as a projection of the state space onto a lower-dimensional space.

Geometric representation of the joint numerical range of observables, illustrating the projection of the state space.

To prove the $5d-7$ bound, they explicitly construct sets of Hermitian matrices using a "diagonal filling" technique, ensuring that any operator in the orthogonal complement of the measurement space has at least two positive and two negative eigenvalues. They also utilize the Hausdorff–Toeplitz theorem and group theory to identify conditions under which UDP automatically implies UDA.

Key Insights: The UDA-UDP Gap and Symmetry

  • The "UDA-UDP Gap": There is a quantifiable gap between the number of observables needed to distinguish a state from other pure states ($4d-5$) versus all states ($5d-7$). This suggests that "proving" purity from data is strictly harder than "identifying" a state known to be pure.
  • Symmetry as a Bridge: One of the most elegant insights is that symmetry can collapse the gap. If the set of measurement observables is the "fixed point" of a certain group of symmetries acting on the state space, then any state that is UDP is also UDA.
  • The Two-Observable Rule: For the simplest case where only two observables are measured, the authors prove that UDP always implies UDA. This is rooted in the inherent convexity of the numerical range for two operators.
  • Multipartite Systems: When dealing with many-body systems, the authors show that "almost all" states can be uniquely determined by their Reduced Density Matrices (RDMs)—essentially, local snapshots of the system—provided the dimensions of the subsystems satisfy specific inequalities (e.g., $d_1 \geq \min(d_2, d_3)$).

Key Results

  • General Bound: For a $d$-dimensional system, there exists a set of $5d - 7$ observables that makes every pure state UDA.
  • Low-Rank Extension: The results extend to low-rank states (states that are "nearly" pure). For a state of rank $q$, roughly $(4q+1)d$ observables are sufficient for UDA.
  • RDM Improvements: For three-particle systems, the authors improved existing bounds, showing that if the first particle's dimension $d_1$ is at least the minimum of the other two, almost all states are UDA by their 2-body RDMs.
  • Symmetry Theorem: Theorem 6 provides a rigorous framework showing that if a compact group of symmetries exists such that the observables are the fixed points, UDP = UDA.

Limitations and Future Work

While $5d-7$ is a significant improvement over $d^2-1$, it is not yet known if it is the absolute minimum; a tighter bound closer to $4d-5$ may exist. Furthermore, while the paper provides a general construction for the $5d-7$ observables, finding "physically natural" observables (like specific Hamiltonians or local interactions) that satisfy these properties for arbitrary $d$ remains an open challenge. Finally, a gap still exists between these UDA results for multipartite systems and known UDP results for RDMs.

Summary updated

Abstract

We discuss the uniqueness of quantum states compatible with given measurement results for a set of observables. For a given pure state, we consider two different types of uniqueness: (1) no other pure state is compatible with the same measurement results and (2) no other state, pure or mixed, is compatible with the same measurement results. For case (1), it was known that for a $d$-dimensional Hilbert space, there exists a set of $4d\ensuremath{-}5$ observables that uniquely determines any pure state. We show that for case (2), $5d\ensuremath{-}7$ observables suffice to uniquely determine any pure state. Thus, there is a gap between the results for (1) and (2), and we give some examples to illustrate this. Unique determination of a pure state by its reduced density matrices (RDMs), a special case of determination by observables, is also discussed. We improve the best-known bound on local dimensions in which almost all pure states are uniquely determined by their RDMs for case (2). We further discuss circumstances where (1) can imply (2). We use convexity of the numerical range of operators to show that when only two observables are measured, (1) always implies (2). More generally, if there is a compact group of symmetries of the state space which has the span of the observables measured as the set of fixed points, then (1) implies (2). We analyze the possible dimensions for the span of such observables. Our results extend naturally to the case of low-rank quantum states.
BibTeX
@inproceedings{chen2013uniqueness,
  author = {Jianxin Chen and Hillary Dawkins and Zhengfeng Ji and Nathaniel Johnston and David W. Kribs and Frederic W. Shultz and Bei Zeng},
  title = {{Uniqueness of quantum states compatible with given measurement results}},
  booktitle = {Physical},
  year = {2013},
  doi = {10.1103/physreva.88.012109},
}
Figure