Preprint
Article

Continuous Deutsch Uncertainty Principle and Continuous Kraus Conjecture

Altmetrics

Downloads

118

Views

121

Comments

0

This version is not peer-reviewed

Submitted:

26 July 2023

Posted:

27 July 2023

You are already at the latest version

Alerts
Abstract
Let $(\Omega, \mu)$, $(\Delta, \nu)$ be measure spaces and $\{\tau_\alpha\}_{\alpha\in \Omega}$, $\{\omega_\beta\}_{\beta \in \Delta}$ be 1-bounded continuous Parseval frames for a Hilbert space $\mathcal{H}$. Then we show that \begin{align}\label{UE} \log (\mu(\Omega)\nu(\Delta))\geq S_\tau(h)+S_\omega (h)\geq -2 \log \left(\frac{1+\displaystyle \sup_{\alpha \in \Omega, \beta \in \Delta}|\langle\tau_\alpha , \omega_\beta\rangle|}{2}\right) , \quad \forall h \in \mathcal{H}_\tau \cap \mathcal{H}_\omega, \end{align} where \begin{align*} &\mathcal{H}_\tau := \{h_1 \in \mathcal{H}: \langle h_1 , \tau_\alpha \rangle \neq 0, \alpha \in \Omega\}, \quad \mathcal{H}_\omega := \{h_2 \in \mathcal{H}: \langle h_2, \omega_\beta \rangle \neq 0, \beta \in \Delta\},\\ &S_\tau(h):= -\displaystyle\int\limits_{\Omega}\left|\left \langle \frac{h}{\|h\|}, \tau_\alpha\right\rangle \right|^2\log \left|\left \langle \frac{h}{\|h\|}, \tau_\alpha\right\rangle \right|^2\,d\mu(\alpha), \quad \forall h \in \mathcal{H}_\tau, \\ & S_\omega (h):= -\displaystyle\int\limits_{\Delta}\left|\left \langle \frac{h}{\|h\|}, \omega_\beta\right\rangle \right|^2\log \left|\left \langle \frac{h}{\|h\|}, \omega_\beta\right\rangle \right|^2\,d\nu(\beta), \quad \forall h \in \mathcal{H}_\omega. \end{align*} We call Inequality (\ref{UE}) as \textbf{Continuous Deutsch Uncertainty Principle}. Inequality (\ref{UE}) improves the uncertainty principle obtained by Deutsch \textit{[Phys. Rev. Lett., 1983]}. We formulate Kraus conjecture for 1-bounded continuous Parseval frames. We also derive continuous Deutsch uncertainty principles for Banach spaces.
Keywords: 
Subject: Computer Science and Mathematics  -   Analysis

1. Introduction

Let H be a finite dimensional Hilbert space. Given an orthonormal basis { ω j } j = 1 n for H , the (finite) Shannon entropy at a point h H τ is defined as
S τ ( h ) j = 1 n h h , τ j 2 log h h , τ j 2 0 ,
where H τ { h H : h , τ j 0 , 1 j n } . In 1983, Deutsch derived following uncertainty principle for Shannon entropy [3].
Theorem 1.1. 
(Deutsch Uncertainty Principle) [3] Let { τ j } j = 1 n , { ω j } j = 1 n be two orthonormal bases for a finite dimensional Hilbert space H . Then
2 log n S τ ( h ) + S ω ( h ) 2 log 1 + max 1 j , k n | τ j , ω k | 2 0 , h H τ H ω .
In 1988, followed by a conjecture of Kraus [9] made in 1987, Maassen and Uffink improved Inequality (1) [12].
Theorem 1.2. (Kraus Conjecture/Maassen-Uffink Uncertainty Principle) [9,12] Let { τ j } j = 1 n , { ω j } j = 1 n be two orthonormal bases for a finite dimensional Hilbert space H . Then
2 log n S τ ( h ) + S ω ( h ) 2 log max 1 j , k n | τ j , ω k | 0 , h H τ H ω .
In 2013, Ricaud and Torrésani [13] showed that Theorem 1.2 holds for Parseval frames.
Theorem 1.3. (Maassen-Uffink-Ricaud-Torrésani Uncertainty Principle) [13] Let { τ j } j = 1 n , { ω j } j = 1 n be two Parseval frames for a finite dimensional Hilbert space H . Then
2 log n S τ ( h ) + S ω ( h ) 2 log max 1 j , k n | τ j , ω k | 0 , h H τ H ω .
Recently, Banach space versions of Deutsch uncertainty principle have been derived in [11]. To formulate them, we need some notions. Given a Parseval p-frame { f j } j = 1 n for X , we define the (finite) p-Shannon entropy at a point x X f as
S f ( x ) j = 1 n f j x x p log f j x x p 0 ,
where X f { x X : f j ( x ) 0 , 1 j n } . On the other way, given a Parseval p-frame { τ j } j = 1 n for X * , we define the (finite) p-Shannon entropy at a point f X τ * as
S τ ( f ) j = 1 n f ( τ j ) f p log f ( τ j ) f p 0 ,
where X τ * { f X * : f ( τ j ) 0 , 1 j n } .
Theorem 1.4. 
[11] (Functional Deutsch Uncertainty Principle) Let { f j } j = 1 n and { g k } k = 1 m be Parseval p-frames for a finite dimensional Banach space X . Then
1 ( n m ) 1 p sup y X , y = 1 max 1 j n , 1 k m | f j ( y ) g k ( y ) |
and
log ( n m ) S f ( x ) + S g ( x ) p log sup y X f X g , y = 1 max 1 j n , 1 k m | f j ( y ) g k ( y ) | > 0 , x X f X g .
Theorem 1.5. 
(Functional Deutsch Uncertainty Principle) Let { τ j } j = 1 n and { ω k } k = 1 m be two Parseval p-frames for the dual X * of a finite dimensional Banach space X . Then
1 ( n m ) 1 p sup g X * , g = 1 max 1 j n , 1 k m | g ( τ j ) g ( ω k ) |
and
log ( n m ) S τ ( f ) + S ω ( f ) p log sup g X τ * X ω * , g = 1 max 1 j n , 1 k m | g ( τ j ) g ( ω k ) | > 0 , f X τ * X ω * .
In this paper, we derive continuous versions of Theorem 1.1, Theorem 1.4 and Theorem 1.5. We also formulate a conjecture based on Theorem 1.2. We wish to say that functional continuous uncertainty principles are derived in [10].

2. Continuous Deutsch Uncertainty Principle and Continuous Kraus Conjecture

In the paper, K denotes C or R and H (resp. X ) denotes a Hilbert space (resp. Banach space) (need not be finite dimensional) over K . We use ( Ω , μ ) to denote a measure space. Continuous frames are introduced independently by Ali, Antoine and Gazeau [1] and Kaiser [8]. In the paper, K denotes C or R and H denotes a finite dimensional Hilbert space.
Definition 2.1. 
[1,8] Let ( Ω , μ ) be a measure space. A collection { τ α } α Ω in a Hilbert space H is said to be a continuous Parseval frame for H if the following conditions hold.
(i) 
For each h H , the map Ω α h , τ α K is measurable.
(ii) 
h 2 = Ω | h , τ α | 2 d μ ( α ) , h H .
We consider the following subclass of continuous Parseval frames.
Definition 2.2. 
A continuous Parseval frame { τ α } α Ω for H is said to be 1-bounded if
τ α 1 , α Ω .
Note that if { τ j } j = 1 n is a Parseval frame for a Hilbert space H , then τ j 1 , 1 j n (see Remark 3.12 in [7]). We are unable to derive this for continuous frames. Given a continuous 1-bounded Parseval frame { τ α } α Ω for H , we define the continuous Shannon entropy at a point h H τ as
S τ ( h ) Ω h h , τ α 2 log h h , τ α 2 d μ ( α ) 0 ,
where H τ { h H : h , τ α 0 , α Ω } . Following is the first fundamental result of this paper.
Theorem 2.3. (Continuous Deutsch Uncertainty Principle) Let ( Ω , μ ) , ( Δ , ν ) be measure spaces and { τ α } α Ω , { ω β } β Δ be 1-bounded continuous Parseval frames for a Hilbert space H . Then
log ( μ ( Ω ) ν ( Δ ) ) S τ ( h ) + S ω ( h ) 2 log 1 + sup α Ω , β Δ | τ α , ω β | 2 0 , h H τ H ω .
Proof. 
Since 1 = Ω h h , τ α 2 d μ ( α ) for all h H { 0 } , 1 = Δ h h , ω β 2 d ν ( β ) for all h H { 0 } and log is concave, using Jensen’s inequality (cf. [6]) we get
S τ ( h ) + S ω ( h ) = Ω h h , τ α 2 log 1 h h , τ α 2 d μ ( α ) + Δ h h , ω β 2 log 1 h h , ω β 2 d ν ( β ) log Ω h h , τ α 2 1 h h , τ α 2 d μ ( α ) + log Δ h h , ω β 2 1 h h , ω β 2 d ν ( β ) = log ( μ ( Ω ) ) + log ν ( Δ ) ) = log ( μ ( Ω ) ν ( Δ ) ) , h H τ H ω .
Let h H τ H ω . Then using Buzano inequality [2,5] we get
S τ ( h ) + S ω ( h ) = Δ Ω h h , τ α 2 h h , ω β 2 log h h , τ α 2 + log h h , ω β 2 d μ ( α ) d ν ( β ) = Δ Ω h h , τ α 2 h h , ω β 2 log h h , τ α h h , ω β 2 d μ ( α ) d ν ( β ) = 2 Δ Ω h h , τ α 2 h h , ω β 2 log h h , τ α h h , ω β d μ ( α ) d ν ( β ) 2 Δ Ω h h , τ α 2 h h , ω β 2 log h h 2 τ α ω β + | τ α , ω β | 2 d μ ( α ) d ν ( β ) = 2 Δ Ω h h , τ α 2 h h , ω β 2 log 1 + | τ α , ω β | 2 d μ ( α ) d ν ( β ) 2 Δ Ω h h , τ α 2 h h , ω β 2 log 1 + sup α Ω , β Δ | τ α , ω β | 2 d μ ( α ) d ν ( β ) = 2 log 1 + sup α Ω , β Δ | τ α , ω β | 2 Δ Ω h h , τ α 2 h h , ω β 2 d μ ( α ) d ν ( β ) 2 log 1 + sup α Ω , β Δ | τ α , ω β | 2 .
Theorem 2.3 promotes following question.
Question 2.4. 
Let ( Ω , μ ) , ( Δ , ν ) be measure spaces, H be a Hilbert space. For which pairs of 1-bounded continuous Parseval frames { τ α } α Ω and { ω β } β Δ for H , we have equality in Inequality (2)?
Based on Theorems 1.3 and Theorem 2.3 we formulate following conjecture.
Conjecture 2.5. (Continuous Kraus Conjecture) Let ( Ω , μ ) , ( Δ , ν ) be measure spaces and { τ α } α Ω , { ω β } β Δ be 1-bounded continuous Parseval frames for a Hilbert space H . Then
S τ ( h ) + S ω ( h ) 2 log sup α Ω , β Δ | τ α , ω β | 0 , h H τ H ω .
Next we derive continuous Deutsch uncertainty for Banach spaces. We need a definition.
Definition 2.6. 
[4] Let ( Ω , μ ) be a measure space and X be a Banach space over K . A collection { f α } α Ω in X * is said to be acontinuous Parseval p-frame( 1 p < ) for X if the following conditions hold.
(i) 
For each x X , the map Ω α f α ( x ) K is measurable.
(ii) 
x p = Ω | f α ( x ) | p d μ ( α ) , x X .
If τ α 1 , α Ω , then we say that the frame { f α } α Ω is 1-bounded.
Similar to Definition 2.2, we set the following.
Definition 2.7. 
A continuous Parseval p-frame { f α } α Ω for X is said to be 1-bounded if
f α 1 , α Ω .
Given a 1-bounded continuous Parseval p-frame { f α } α Ω for X , we define the continuous p-Shannon entropy at a point x X f as
S f ( x ) Ω f α x x p log f α x x p d μ ( α ) 0 ,
where X f { x X : f α ( x ) 0 , α Ω } .
Theorem 2.8. (Functional Continuous Deutsch Uncertainty Principle ) Let ( Ω , μ ) , ( Δ , ν ) be measure spaces and { f α } α Ω , { g β } β Δ be 1-bounded continuous Parseval p-frames for a Banach space X . Then
1 ( μ ( Ω ) ν ( Δ ) ) 1 p sup y X , y = 1 sup α Ω , β Δ | f α ( y ) g β ( y ) |
and
S f ( x ) + S g ( x ) p log sup y X , y = 1 sup α Ω , β Δ | f α ( y ) g β ( y ) | 0 , x X f X g .
Proof. 
Let z X be such that z = 1 . Then
1 = Ω | f α ( z ) | p d μ ( α ) Δ | g β ( z ) | p d ν ( β ) = Ω Δ | f α ( z ) g β ( z ) | p d ν ( β ) d μ ( α ) Ω Δ sup y X , y = 1 sup α Ω , β Δ | f α ( y ) g β ( y ) | p d ν ( β ) d μ ( α ) = sup y X , y = 1 sup α Ω , β Δ | f α ( y ) g β ( y ) | p μ ( Ω ) ν ( Δ )
which gives
1 μ ( Ω ) ν ( Δ ) sup y X , y = 1 sup α Ω , β Δ | f α ( y ) g β ( y ) | p .
Let x X f X g . Then
S f ( x ) + S g ( x ) = Ω Δ f α x x p g β x x p log f α x x p + log g β x x p d ν ( β ) d μ ( α ) = Ω Δ f α x x p g β x x p log f α x x g β x x p d ν ( β ) d μ ( α ) = p Ω Δ f α x x p g β x x p log f α x x g β x x d ν ( β ) d μ ( α ) p Ω Δ f α x x p g β x x p log sup y X , y = 1 sup α Ω , β Δ | f α ( y ) g β ( y ) | d ν ( β ) d μ ( α ) = p log sup y X , y = 1 sup α Ω , β Δ | f α ( y ) g β ( y ) | Ω Δ f α x x p g β x x p d ν ( β ) d μ ( α ) = p log sup y X , y = 1 sup α Ω , β Δ | f α ( y ) g β ( y ) | .
Corollary 2.9. 
Theorem 1.1 follows from Theorem 2.8.
Proof. 
Let ( Ω , μ ) , ( Δ , ν ) be measure spaces and { τ α } α Ω , { ω β } β Δ be 1-bounded continuous Parseval frames for a Hilbert space H . Define
f α : H h h , τ α K ; α Ω , g β : H h h , ω β K , β Δ .
Now by using Buzano inequality [2,5] we get
sup h H , h = 1 sup α Ω , β Δ | f α ( h ) g β ( h ) | = sup h H , h = 1 sup α Ω , β Δ | h , τ α | | h , ω β | sup h H , h = 1 sup α Ω , β Δ h 2 τ α ω β + | τ α , ω β | 2 = 1 + sup α Ω , β Δ | τ j , ω k | 2 .
Theorem 2.8 brings the following question.
Question 2.10. 
Let ( Ω , μ ) , ( Δ , ν ) be measure spaces, X be a Banach space and p > 1 . For which pairs of continuous Parseval p-frames { f α } α Ω , and { g β } β Δ for X , we have equality in Inequality (4)?
Next we derive a dual inequality of (4). For this we need dual of Definition 2.6.
Definition 2.6. 
Let ( Ω , μ ) be a measure space and X be a Banach space over K . A collection { τ α } α Ω in X is said to be a continuous Parseval p-frame ( 1 p < ) for X * if the following conditions hold.
(i) 
For each ϕ X * , the map Ω α ϕ ( τ α ) K is measurable.
(ii) 
f p = Ω | f ( τ α ) | p d μ ( α ) , f X * .
f τ α 1 , α Ω , then we say that the frame { τ α } α Ω is 1-bounded.
Given a 1-bounded continuous Parseval p-frame { τ α } α Ω for X * , we define the continuous p-Shannon entropy at a point f X τ * as
S f ( x ) Ω f ( τ α ) f p log f ( τ α ) f p d μ ( α ) 0 ,
where X τ * { f X : f ( τ α ) 0 , α Ω } } . We now have the following dual to Theorem 2.8.
Theorem .(Continuous Deutsch Uncertainty Principle for Banach spaces) Let ( Ω , μ ) , ( Δ , ν ) be measure spaces and { τ α } α Ω , { ω β } β Δ be 1-bounded continuous Parseval p-frames for the dual X * of a Banach space X . Then
1 ( μ ( Ω ) ν ( Δ ) ) 1 p sup g X * , g = 1 sup α Ω , β Δ | g ( τ α ) g ( ω β ) |
and
S τ ( f ) + S ω ( f ) p log sup g X * , g = 1 sup α Ω , β Δ | g ( τ α ) g ( ω β ) | 0 , f X τ * X ω * .
Proof. 
Let h X * be such that h = 1 . Then
1 = Ω | h ( τ α ) | p d μ ( α ) Δ h ( ω β ) | p d ν ( β ) = Ω Δ | h ( τ α ) h ( ω β ) | p d ν ( β ) d μ ( α ) Ω Δ sup g X * , g = 1 sup α Ω , β Δ | g ( τ α ) g ( ω β ) | p d ν ( β ) d μ ( α ) = sup g X * , g = 1 sup α Ω , β Δ | g ( τ α ) g ( ω β ) | p μ ( Ω ) ν ( Δ )
which gives
1 μ ( Ω ) ν ( Δ ) sup g X * , g = 1 sup α Ω , β Δ | g ( τ α ) g ( ω β ) | p .
Let f X τ * X ω * . Then
S τ ( f ) + S ω ( f ) = Ω Δ f ( τ α ) f p f ( ω β ) f p log f ( τ α ) f p + log f ( ω β ) f p d ν ( β ) d μ ( α ) = Ω Δ f ( τ α ) f p f ( ω β ) f p log f ( τ α ) f f ( ω β ) f p d ν ( β ) d μ ( α ) = p Ω Δ f ( τ α ) f p f ( ω β ) f p log f ( τ α ) f f ( ω β ) f d ν ( β ) d μ ( α ) p Ω Δ f ( τ α ) f p f ( ω β ) f p log sup g X * , g = 1 sup α Ω , β Δ | g ( τ α ) g ( ω β ) | d ν ( β ) d μ ( α ) = p log sup g X * , g = 1 sup α Ω , β Δ | g ( τ α ) g ( ω β ) | Ω Δ f ( τ α ) f p f ( ω β ) f p d ν ( β ) d μ ( α ) = p log sup g X * , g = 1 sup α Ω , β Δ | g ( τ α ) g ( ω β ) | .
Theorem 2.12 again gives the following question.
Question 2.13. 
Let ( Ω , μ ) , ( Δ , ν ) be measure spaces, X be a Banach space and p > 1 . For which pairs of continuous Parseval p-frames { τ α } α Ω , and { ω β } β Δ for X * , we have equality in Inequality (5)

References

  1. S. Twareque Ali, J.-P. Antoine, and J.-P. Gazeau. Continuous frames in Hilbert space. Ann. Physics, 222(1):1–37, 1993.
  2. Maria Luisa Buzano. Generalizzazione della diseguaglianza di Cauchy-Schwarz. Rend. Sem. Mat. Univ. e Politec. Torino, 31:405–409 (1974), 1971/73.
  3. David Deutsch. Uncertainty in quantum measurements. Phys. Rev. Lett., 50(9):631–633, 1983.
  4. Mohammad Hasan Faroughi and Elnaz Osgooei. Continuous p-Bessel mappings and continuous p-frames in Banach spaces. Involve, 4(2):167–186, 2011.
  5. Masatoshi Fujii and Fumio Kubo. Buzano’s inequality and bounds for roots of algebraic equations. Proc. Amer. Math. Soc., 117(2):359–361, 1993.
  6. D. J. H. Garling. Inequalities: a journey into linear analysis. Cambridge University Press, Cambridge, 2007.
  7. Deguang Han, Keri Kornelson, David Larson, and Eric Weber. Frames for undergraduates, volume 40 of Student Mathematical Library. American Mathematical Society, Providence, RI, 2007.
  8. Gerald Kaiser. A friendly guide to wavelets. Modern Birkhäuser Classics. Birkhäuser/Springer, New York, 2011.
  9. K. Kraus. Complementary observables and uncertainty relations. Phys. Rev. D (3), 35(10):3070–3075, 1987.
  10. K. Mahesh Krishna. Functional continuous uncertainty principle. Preprints:10.20944/preprints202307.1643.v1, 25 July, 2023.
  11. K. Mahesh Krishna. Functional Deutsch uncertainty principle. Preprints:10.20944/preprints202307.1084.v1, 17 July, 2023.
  12. Hans Maassen and J. B. M. Uffink. Generalized entropic uncertainty relations. Phys. Rev. Lett., 60(12):1103–1106, 1988.
  13. Benjamin Ricaud and Bruno Torrésani. Refined support and entropic uncertainty inequalities. IEEE Trans. Inform. Theory, 59(7):4272–4279, 2013.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

Subscribe

© 2024 MDPI (Basel, Switzerland) unless otherwise stated