The uncertainty geometry of finite-dimensional position and momentum

Dimpi Thakuria, Shuheng Liu, Giuseppe Vitagliano, Konrad Szymański · 2026

Read The uncertainty geometry of finite-dimensional position and momentum by Dimpi Thakuria, Shuheng Liu, Giuseppe Vitagliano, Konrad Szymański (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-11876.txt.

Introduction

Dimpi Thakuria and three colleagues map which covariance matrices a finite position-momentum pair can attain. The pair is linked by the discrete Fourier transform. They combine pen-and-paper math with convex shapes and matrix methods over joint number ranges. They describe the allowed region through unitary markers such as trace and determinant. This pins down edge states, extends states of least uncertainty, and yields bounds for multi-knob sensing and tests for whether two-part systems split apart.

Catalog record

CREATORDimpi Thakuria, Shuheng Liu, Giuseppe Vitagliano, Konrad Szymański
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE80716
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.11876

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The uncertainty geometry of finite-dimensional position and momentum Dimpi Thakuria, Shuheng Liu, Giuseppe Vitagliano, Konrad Szymański Abstract Uncertainty relations are usually stated as bounds on selected combinations of variances, but the full covariance matrix contains substantially richer information about the geometry of quantum state space and about the operational capabilities of quantum systems. Here we characterize the covariance matrices attainable by a finite-dimensional canonical pair of observables related by the discrete Fourier transform, the natural analogue of position and momentum in a finite Hilbert space. We combine analytic arguments with convex-geometric and semidefinite-programming methods based on joint numerical ranges to describe the admissible region through unitary invariants, in particular the trace and determinant of the covariance matrix. This provides a systematic way to identify extremal states, generalizing the notion of minimum-uncertainty states, and to quantify how the discrete uncertainty geometry approaches its continuous counterpart with increasing dimension. We further show that the resulting covariance-matrix characterization has direct consequences for applications: it yields accuracy bounds for multiparameter estimation protocols and separability criteria for finite-dimensional bipartite systems, including discrete analogues of continuous-variable EPR-type witnesses. Our results establish a systematic and versatile platform for connecting uncertainty relations, convex quantum geometry, metrology, and entanglement detection in finite-dimensional systems. I Introduction Uncertainty relations are among the most distinctive structural features of quantum theory. In their most common formulations, they quantify the impossibility of preparing states with simultaneously sharp values for incompatible observables, and they are most naturally expressed in terms of second moments and covariance matrices 69 ; 70 . Over the years, this perspective has been complemented by many alternative formulations, including entropic uncertainty relations 98 ; 16 , characteristic-function approaches 72 , and refinements involving monotone metrics and quantum Fisher information 15 ; 63 ; 33 ; 4 ; 86 . Historically, uncertainty relations were already central to the discussion of completeness in quantum mechanics initiated by Einstein, Podolsky, and Rosen and by Bohr’s reply 27 ; 9 . Their utility is not restricted to historical or foundational arguments, though 101 : in modern quantum information, covariance matrices and uncertainty relations provide powerful tools that find applications ranging from quantum metrology 85 to entanglement detection and quantification 39 ; 99 ; 43 ; 37 ; 35 ; 36 ; 52 . For example, states that minimize variance-based uncertainty relations are intimately related to coherent and squeezed states 44 ; 21 ; 90 ; 89 ; 48 ; 102 , which play a central role in quantum optics and in precision measurements beyond the standard quantum limit, and, in many-particle systems they are entangled 80 ; 81 ; 85 ; 65 ; 30 . More generally, the covariance matrix is closely tied to metrological sensitivity through its relation to the quantum Fisher information 66 ; 63 ; 61 ; 103 ; 87 ; 51 . The properties of quantum mechanics captured by uncertainty relations have led to the development of a rich mathematical subject, not restricted to lower bounds and sum of variances only. This includes geometric descriptions of uncertainty regions and of the extremal states that bound them 46 ; 47 , as well as state-independent and many-observable generalizations 20 ; 19 ; 73 ; 79 . A natural next step is to characterize the full set of achievable covariance matrices, and in particular its boundary, capturing the extremal variance behavior 49 ; 62 ; 50 . The most paradigmatic uncertainty relations often involve canonical position/momentum pairs, for historical reasons but also for their maximal incompatibility 11 . In fact, complementary approaches focused on the measurement uncertainty, capturing the properties of joint approximate measurements are also centered mostly on position and momentum being the canonical example 100 ; 12 ; 13 . We focus on the covariance-matrix viewpoint, directly defined by states and observables, within the framework of preparation uncertainty , connecting algebraic incompatibility, geometry of state space, and experimentally accessible second moments. Specifically, we focus on the finite-dimensional analogue of a canonical pair, which also in some sense share the maximal incompatibility property. These operators are obtained from one another through the finite Fourier transform, a standard construction in discrete phase-space methods 93 ; 95 , which in particular also implies that the eigenbases of the two operators are maximally unbiased 26 . Finite Fourier-related observables have appeared in several forms in the literature on discrete phase spaces and discrete uncertainty relations 56 ; 20 , but a systematic covariance-matrix analysis of their jointly achievable second moments is, to our knowledge, still missing. Understanding this problem is interesting both conceptually and practically. Conceptually, finite-dimensional Fourier pairs provide a controlled setting in which one can compare discrete and continuous canonical structures, which is a long-standing and still active subject of intense investigation, including definition of finite phase space 78 ; 3 ; 93 ; 104 ; 94 ; 5 ; 7 ; 18 ; 53 ; 2 ; 95 as well as finite-dimensional analogue of the harmonic oscillator 8 ; 18 ; 2 . Practically, they are relevant whenever continuous quadratures are accessed only through finite resolution or coarse-grained measurements, as happens in realistic detection schemes; in such scenarios, complementary coarse-grained observables and their uncertainty relations become operationally meaningful 71 ; 84 ; 83 ; 60 ; 82 . The shift from continuous variables provides a richer geometry, too: finite-dimensional operators have bounded spectra, unlike their continous variable counterparts, leading to compactness of the achievable covariance matrices, and the existence of nontrivial upper bounds to uncertainty, completely absent in the CV scenario. Our main goal is to fully describe the admissible preparation uncertainty region of a finite-dimensional canonical pair, by characterizing the boundary of the set of allowed covariance matrices. In particular, projecting the region of admissible covariance matrices into natural invariants such as trace and determinant we identify states corresponding to the boundary of the uncertainty region, i.e., extremal uncertainty states. These provide also a generalized set of states that are useful for applications, especially in metrology. In this way, we obtain a finite-dimensional uncertainty geometry that parallels several well-known features of the continuous-variable case while also displaying genuinely discrete effects. This parallel also extends to the multipartite case: The characterization of the allowed covariance matrices of a single particle allows us to find entanglement criteria which provide a finite-dimensional analogue of the covariance matrix criterion, which is of paramount importance in CV systems as it allows to fully characterize and quantify entanglement of multi-mode Gaussian states 1 ; 97 . The paper is organized as follows. First we introduce some background methods: In section I.1 , we review the basic properties of covariance matrices and their connection to the Robertson-Schrödinger uncertainty relation, while in section I.2 , we introduce the finite-dimensional analogues of position and momentum operators, constructed via the finite Fourier transform. Afterwards, we present our results: In section II , we describe numerical tools for characterizing the set of achievable covariance matrices, including joint numerical range methods and sampling algorithms applied to the canonical pair. In section III , we discuss applications to metrology and entanglement. Focusing on the multiparameter estimation of finite-dimensional displacement operators we firstly derive accuracy bounds coming from the quantum Fisher information matrix in section

Read the complete text