Tight Generalization of Robertson-Type Uncertainty Relations

Gen Kimura, Aina Mayumi, Haruki Yamashita · 2025

Read Tight Generalization of Robertson-Type Uncertainty Relations by Gen Kimura, Aina Mayumi, Haruki Yamashita (2025) 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-2505-19861.txt.

Introduction

Gen Kimura, Aina Mayumi, and Haruki Yamashita tighten a quantum uncertainty bound in a 2025 paper. Uncertainty rules cap how jointly two measures can be pinned down. The classic Robertson form uses a fixed constant that ignores the state at hand.

Their new constant turns on the state's largest and smallest weights, and they prove no sharper form of this type exists. All cases that hit the bound exactly are also pinned down. The gain grows as the state grows more mixed, a trade-off the old rule missed. The result also sharpens error-disturbance bounds for measuring devices.

Catalog record

CREATORGen Kimura, Aina Mayumi, Haruki Yamashita
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-19861

Browse the full library · About the Tower of Babel

Full text

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

Tight Generalization of Robertson-Type Uncertainty Relations Gen Kimura, Aina Mayumi, Haruki Yamashita Abstract We establish the tightest possible Robertson-type preparation uncertainty relation, which explicitly depends on the eigenvalues of the quantum state. The conventional constant 1 4 \tfrac{1}{4} is replaced by a state-dependent coefficient ( λ max + λ min ) 2 4 ​ ( λ max − λ min ) 2 , \frac{(\lambda_{\max}+\lambda_{\min})^{2}}{4(\lambda_{\max}-\lambda_{\min})^{2}}, where λ max \lambda_{\max} and λ min \lambda_{\min} denote the largest and smallest eigenvalues of the density operator ρ \rho , respectively. This coefficient is optimal among all Robertson-type generalizations and does not admit further improvement. Our relation becomes more pronounced as the quantum state becomes more mixed, capturing a trade-off in quantum uncertainty that the conventional Robertson’s relation fails to detect. In addition, our result also provides a strict generalization of the Schrödinger’s uncertainty relation, showing that the uncertainty trade-off is governed by the sum of the covariance term and a state-dependent improvement over the Robertson bound. As applications, we also refine error–disturbance trade-offs by incorporating spectral information of both the system and the measuring apparatus, thereby generalizing the Arthurs–Goodman and Ozawa inequalities. I Introduction The canonical uncertainty principle reveals that two non‑commuting observables cannot simultaneously assume arbitrarily sharp values [ 1 , 2 , 3 , 4 ] . Building on Kennard’s formulation of uncertainty in terms of standard deviations for position and momentum observables [ 5 ] , Robertson’s uncertainty relation [ 6 ] established a general trade-off relation between arbitrary observables, thereby revealing an intrinsic quantum uncertainty: \displaystyle V_{\rho}(A)V_{\rho}(B) \displaystyle\geq\frac{1}{4}\Bigl|\langle[A,B]\rangle_{\rho}\Bigr|^{2}, (1) where A , B are observables represented by Hermitian operators, V_{\rho}(X):=\Tr[(X-\langle X\rangle_{\rho})^{2}\rho] and \langle X\rangle_{\rho}:=\Tr[X\rho] denote the variance of and the expectation value of X in the quantum state \rho , and [A,B]:=AB-BA denotes the commutator. The relation was soon refined by Schrödinger [ 7 ] , who added to the Robertson bound a quantum covariance term \mathrm{Cov}_{\rho}(A,B)\coloneqq\frac{1}{2}\langle\{A,B\}\rangle_{\rho}-\langle A\rangle_{\rho}\langle B\rangle_{\rho}, with the anticommutator \{A,B\}\coloneqq AB+BA : V_{\rho}(A)\,V_{\rho}(B)\geq\frac{1}{4}\Bigl|\langle[A,B]\rangle_{\rho}\Bigr|^{2}+\mathrm{Cov}_{\rho}(A,B)^{2}. (2) Following the pioneering work of Robertson and its refinement by Schrödinger, many developments have been made on preparation-type uncertainty relations. Maccone and Pati [ 8 ] proposed an improvement to Robertson’s relation using the sum of variances [ 9 ] , which has led to further developments [ 10 , 11 , 12 ] . On the other hand, Luo [ 13 ] and Park [ 14 ] independently proposed improvements for mixed states using the Wigner–Yanase skew information [ 15 ] , and several related studies have also been conducted [ 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 ] . Another refinement of Robertson’s relation ( 1 ) for mixed state was also proposed by Nagaoka and Hayashi [ 24 , 25 ] , while several studies have characterized uncertainty regions [ 26 , 27 , 28 , 29 ] . Besides variance, formulations using other measures—such as entropy [ 30 , 31 , 32 , 33 , 34 , 35 ] , maximum probabilities [ 36 , 37 ] , Fisher information [ 38 , 39 , 40 ] , and quantum coherence [ 41 , 42 , 43 , 44 , 45 , 46 ] —have also been developed. In a recent work [ 47 ] , we considered the following type of uncertainty relation: V_{\rho}(A)\,V_{\rho}(B)\geq c(\rho)\bigl\|[A,B]\bigr\|^{2}_{\rho}, (3) where \|X\|_{\rho}^{2}:=\Tr( \rho X^\dagger X ) is the state-dependent norm, which generalizes the Frobenius (Hilbert–Schmidt) norm, and c(\rho) is a positive constant that may depend on the state \rho . We showed that c(\rho)=\lambda_{\rm min}/(2\lambda_{\rm max}^{2}) yields a valid general uncertainty relation, where \lambda_{\rm min} and \lambda_{\rm max} denote the smallest and largest eigenvalues of \rho , respectively, while the optimal bound was conjectured to be c_{\rm opt}(\rho)=\frac{\lambda_{\rm min}\lambda_{\rm Smin}}{\lambda_{\rm min}+\lambda_{\rm Smin}}, (4) where \lambda_{\rm Smin} is the second-smallest eigenvalue of \rho . Recently, in Ref. [ 48 ] , we not only provided a rigorous proof of this result, but also further strengthened the Schrödinger uncertainty relation by incorporating the additional bound ( 4 ) into the original expression ( 2 ): \displaystyle V_{\rho}(A)\,V_{\rho}(B)\geq\ \frac{1}{4}\left|\langle[A,B]\rangle_{\rho}\right|^{2}+\mathrm{Cov}_{\rho}(A,B)^{2} \displaystyle\quad+\frac{\lambda_{\rm min}\lambda_{\rm Smin}}{\lambda_{\rm min}+\lambda_{\rm Smin}}\|[A,B]\|_{\rho}^{2}. (5) The last term reveals a previously overlooked quantum contribution to the uncertainty trade-off, whose significance increases as the quantum state becomes more mixed. Thus, the uncertainty relation of the form ( 3 ), based on the commutator norm dependent on \rho , has been established in its optimal form. It is, however, still of interest to pursue a direct refinement of the Robertson-type relation: Namely, we consider the problem of determining the largest possible positive coefficient c^{\prime}(\rho) such that the following uncertainty relation holds: V_{\rho}(A)V_{\rho}(B)\geq c^{\prime}(\rho)\Bigl|\langle[A,B]\rangle_{\rho}\Bigr|^{2}. (6) Note that the case c^{\prime}(\rho)=1/4 corresponds to the original Robertson’s relation ( 1 ). Somewhat surprisingly, however, this bound is not yet optimal and can be further improved. We show that the optimal bound is determined by the minimum and maximum eigenvalues of the density operator, and is given by c^{\prime}_{\rm opt}(\rho)=\frac{(\lambda_{\rm max}+\lambda_{\rm min})^{2}}{4(\lambda_{\rm max}-\lambda_{\rm min})^{2}}, (7) leading to a direct generalization of Robertson’s uncertainty relation: V_{\rho}(A)\,V_{\rho}(B)\geq\frac{(\lambda_{\rm max}+\lambda_{\rm min})^{2}}{4(\lambda_{\rm max}-\lambda_{\rm min})^{2}}\Bigl|\langle[A,B]\rangle_{\rho}\Bigr|^{2}. (8) This relation yields a strictly sharper bound than the Robertson bound ( 1 ) for all faithful states, i.e., when \lambda_{\min}>0 . Interestingly, while the smallest and the second smallest eigenvalues of \rho determine the optimal bound ( 4 ) in the case of the commutator norm, the optimal bound of the Robertson type is determined by the smallest and the largest eigenvalues of \rho . In the case of the maximally mixed state, \rho_{\max}=\mathop{\mathbb{I}}\nolimits/d for a d -level system, the eigenvalues satisfy \lambda_{\rm max}=\lambda_{\rm min}=1/d , and hence the bound ( 7 ) formally diverges. However, since \left|\langle[A,B]\rangle_{\rho_{\max}}\right|^{2}=0 for the maximally mixed state due to the cyclicity of the trace, the trade-off relation remains nontrivial and continues to hold in a meaningful way. Indeed, by considering a series of faithful state converging to the maximally mixed state, we obtain the following relation: V_{\rho_{\max}}(A)V_{\rho_{\max}}(B)\geq\frac{1}{d^{2}}\|[A,B]\|_{\rm{op}}^{2}, (9) where \|X\|_{\rm{op}}:=\max_{\ket{\psi}\neq 0}\|X\psi\|/\|\psi\| denotes the operator norm. Moreover, using a technique introduced in Ref. [ 48 ] , uncertainty relation ( 8 ) can be further generalized to V_{\rho}(A)V_{\rho}(B)\geq\frac{(\lambda_{\rm max}+\lambda_{\rm min})^{2}}{4(\lambda_{\rm max}-\lambda_{\rm min})^{2}}\left|\langle[A,B]\rangle_{\rho}\right|^{2}+{\rm Cov}_{\rho}(A,B)^{2}, (10) which provides another refinement of the Schrödinger’s uncertainty relation. While preparation-type uncertainty relations have been the focus so far, measurement-type counterparts also play a central role in the understanding of quantum measurement, as they explicitly characterize the trade-off between measurement precision and the disturbance it inevitably induces [ 49 , 50 , 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 ] . As an application of our uncertainty relation, we also obtain generalizations of measurement-type uncertainty relations, including the Arthurs–Goodman inequality and Ozawa’s inequality [ 50 , 51 , 52 ] . The paper is organized as follows. In Sec. II , we discuss a Robertson-type uncertainty relation and derive its tight generalization. In Sec. III , we use this result to obtain a tight generalization of Schrödinger’s relation. In Sec. IV , we compare our generalized bounds with the original Robertson and Schrödinger relations in the qubit case by averaging the bounds over all pairs of observables. In Sec. V , we apply our generalized

Read the complete text