Strong Subadditivity of Quantum Entropy

Strong subadditivity of entropy (SSA) was long known and appreciated in classical probability theory and information theory. Its extension to quantum mechanical entropy (the von Neumann entropy) was conjectured by D.W. Robinson and D. Ruelle ^[1] in 1966 and O. E. Lanford III and D. W. Robinson ^[2] in 1968 and proved in 1973 by E.H. Lieb and M.B. Ruskai.^[3] It is a basic theorem in modern quantum information theory. Renato Renner and Omar Fawzi proved a strengthening of strong subadditivity in 2014 ^[4]

SSA concerns the relation between the entropies of various subsystems of a larger system consisting of three subsystems (or of one system with three degrees of freedom). The proof of this relation in the classical case is quite easy but the quantum case is difficult because of the non-commutativity of the density matrices describing the subsystems.

Some useful references here are.^[5][6][7]

Definitions

We will use the following notation throughout: A Hilbert space is denoted by $\mathcal{H}$, and $\mathcal{B}(\mathcal{H})$ denotes the bounded linear operators on $\mathcal{H}$. Tensor products are denoted by superscripts, e.g., $\mathcal{H}^{12}=\mathcal{H}^1\otimes \mathcal{H}^2$. The trace is denoted by ${\rm Tr}$.

Density matrix

A density matrix is a Hermitian, positive semi-definite matrix of trace one. It allows for the description of a quantum system in a mixed state. Density matrices on a tensor product are denoted by superscripts, e.g., $\rho^{12}$ is a density matrix on $\mathcal{H}^{12}$.

Entropy

The von Neumann quantum entropy of a density matrix $\rho$ is

$S(\rho):=-{\rm Tr}(\rho\log \rho)$.

Relative entropy

Umegaki's^[8] quantum relative entropy of two density matrices $\rho$ and $\sigma$ is

$S(\rho||\sigma)={\rm Tr}(\rho\log\rho-\rho\log\sigma)\geq 0$.

Joint concavity

A function $g$ of two variables is said to be jointly concave if for any $0\leq \lambda\leq 1$ the following holds

$g(\lambda A_1 + (1-\lambda)A_2,\lambda B_1 + (1-\lambda)B_2 ) \geq \lambda g(A_1, B_1) + (1 -\lambda)g(A_2, B_2).$

Subadditivity of entropy

Ordinary subadditivity ^[9] concerns only two spaces $\mathcal{H}^{12}$ and a density matrix $\rho^{12}$. It states that

$S(\rho^{12}) \leq S(\rho^1) +S(\rho^2)$

This inequality is true, of course, in classical probability theory, but the latter also contains the theorem that the conditional entropies $S(\rho^{12} | \rho^1)= S(\rho^{12} )-S(\rho^1)$ and $S(\rho^{12} | \rho^2)=S(\rho^{12} ) -S(\rho^2)$ are both non-negative. In the quantum case, however, both can be negative, e.g. $S(\rho^{12})$ can be zero while $S(\rho^1) = S(\rho^{2}) >0$. Nevertheless, the subadditivity upper bound on $S(\rho^{12})$ continues to hold. The closest thing one has to $S(\rho^{12})- S(\rho^1)\geq 0$ is the Araki–Lieb triangle inequality ^[9]

$S(\rho^{12}) \geq |S(\rho^1) -S(\rho^2)|$

which is derived in ^[9] from subadditivity by a mathematical technique known as 'purification'.

Strong subadditivity (SSA)

Suppose that the Hilbert space of the system is a tensor product of three spaces: $\mathcal{H}=\mathcal{H}^1\otimes \mathcal{H}^2\otimes \mathcal{H}^3.$. Physically, these three spaces can be interpreted as the space of three different systems, or else as three parts or three degrees of freedom of one physical system.

Given a density matrix $\rho^{123}$ on $\mathcal{H}$, we define a density matrix $\rho^{12}$ on $\mathcal{H}^1\otimes \mathcal{H}^2$ as a partial trace: $\rho^{12}={\rm Tr}_{\mathcal{H}^3} \rho^{123}$. Similarly, we can define density matrices: $\rho^{23}$, $\rho^{13}$, $\rho^1$, $\rho^2$, $\rho^3$.

Statement

For any tri-partite state $\rho^{123}$ the following holds

$S(\rho^{123})+S(\rho^2)\leq S(\rho^{12})+S(\rho^{23})$,

where $S(\rho^{12})=-{\rm Tr}_{\mathcal{H}^{12}} \rho^{12} \log \rho^{12}$, for example.

Equivalently, the statement can be recast in terms of conditional entropies to show that for tripartite state $\rho^{ABC}$,

$S(A\mid BC)\leq S(A\mid B)$.

This can also be restated in terms of quantum mutual information,

$I(A:BC)\geq I(A:B)$.

These statements run parallel to classical intuition, except that quantum conditional entropies can be negative, and quantum mutual informations can exceed the classical bound of the marginal entropy.

The strong subadditivity inequality was improved in the following way by Carlen and Lieb ^[10]

$S(\rho^{12})+S(\rho^{23})-S(\rho^{123})-S(\rho^2) \geq 2\max\{S(\rho^1)-S(\rho^{12}),S(\rho^2)-S(\rho^{12}), 0 \}$,

with the optimal constant $2$.

As mentioned above, SSA was first proved by E.H.Lieb and M.B.Ruskai in,^[3] using Lieb's theorem that was proved in.^[11] The extension from a Hilbert space setting to a von Neumann algebra setting, where states are not given by density matrices, was done by Narnhofer and Thirring .^[12]

The theorem can also be obtained by proving numerous equivalent statements, some of which are summarized below.

Wigner–Yanase–Dyson conjecture

E. P. Wigner and M. M. Yanase ^[13] proposed a different definition of entropy, which was generalized by F.J. Dyson.

The Wigner–Yanase–Dyson p-skew information

The Wigner–Yanase–Dyson $p$-skew information of a density matrix $\rho$. with respect to an operator $K$ is

$I_p(\rho, K)=\frac{1}{2}{\rm Tr}[\rho^p, K^*][\rho^{1-p}, K],$

where $[A,B]=AB-BA$ is a commutator, $K^*$ is the adjoint of $K$ and $0\leq p\leq 1$ is fixed.

Concavity of p-skew information

It was conjectured by E. P. Wigner and M. M. Yanase in ^[14] that $p$- skew information is concave as a function of a density matrix $\rho$ for a fixed $0\leq p\leq 1$.

Since the term $-\tfrac{1}{2}{\rm Tr}\rho KK^*$ is concave (it is linear), the conjecture reduces to the problem of concavity of $Tr\rho^p K^*\rho^{1-p}K$. As noted in,^[11] this conjecture (for all $0 \leq p \leq 1$) implies SSA, and was proved for $p= \tfrac{1}{2}$ in,^[14] and for all $0\leq p \leq 1$ in ^[11] in the following more general form: The function of two matrix variables

$A, B \mapsto {\rm Tr} A^{r}K^*B^pK$

(1)

is jointly concave in $A$ and $B,$ when $0\leq r\leq 1$ and $p+r \leq 1$.

This theorem is an essential part of the proof of SSA in.^[3]

In their paper ^[14] E. P. Wigner and M. M. Yanase also conjectured the subadditivity of $p$-skew information for $p=\tfrac{1}{2}$, which was disproved by Hansen^[15] by giving a counterexample.

First two statements equivalent to SSA

It was pointed out in ^[9] that the first statement below is equivalent to SSA and A. Ulhmann in ^[16] showed the equivalence between the second statement below and SSA.

• $S(\rho^1)+S(\rho^3)-S(\rho^{12})-S(\rho^{23})\leq 0.$ Note that the conditional entropies $S(\rho^{12}|\rho^1)$ and $S(\rho^{23}|\rho^3)$ do not have to be both non-negative.
• The map $\rho^{12}\mapsto S(\rho^1)-S(\rho^{12})$ is convex.

Both of these statements were proved directly in.^[3]

Joint convexity of relative entropy

As noted by Lindblad ^[17] and Uhlmann ,^[18] if, in equation (1), one takes $K=1$ and $r=1-p, A=\rho$ and $B=\sigma$ and differentiates in $p$ at $p=0$ one obtains the Joint convexity of relative entropy : i.e., if $\rho=\sum_k\lambda_k\rho_k$, and $\sigma=\sum_k\lambda_k\sigma_k$, then

$S\Bigl(\sum_k \lambda_k\rho_k||\sum_k\lambda_k \sigma_k \Bigr)\leq \sum_k\lambda_k S(\rho_k||\sigma_k),$

(2)

where $\lambda_k\geq 0$ with $\sum_k\lambda_k=1$.

Monotonicity of quantum relative entropy

The relative entropy decreases monotonically under certain operations on density matrices, the most important and basic of which is the following. Consider the map $T$ from $\mathcal{B}(\mathcal{H}^{12}) \rightarrow \mathcal{B}(\mathcal{H}^{12})$ given by $T=1_{\mathcal{H}^1}\otimes Tr_{\mathcal{H}^2}$ . Then

$S(T\rho||T\sigma)\leq S(\rho||\sigma),$

(3)

which is called Monotonicity of quantum relative entropy under partial trace.

To see how this follows from the joint convexity of relative entropy, observe that $T$ can be written in Uhlmann's representation as

$T(\rho^{12} ) = N^{-1} \sum_{j=1}^N (1_{\mathcal{H}^1}\otimes U_j) \rho^{12}(1_{\mathcal{H}^1}\otimes U_j^*),$

for some finite $N$ and some collection of unitary matrices on $\mathcal{H}^2$ (alternatively, integrate over Haar measure). Since the trace (and hence the relative entropy) is unitarily invariant, inequality (3) now follows from (2). This theorem is due to Lindblad ^[17] and Uhlmann,^[16] whose proof is the one given here.

SSA is obtained from (3) with $\mathcal{H}^1$ replaced by $\mathcal{H}^{12}$ and $\mathcal{H}^2$ replaced $\mathcal{H}^3$. Take $\rho = \rho^{123}, \sigma = \rho^1\otimes \rho^{23}, T= 1_{\mathcal{H}^{12}}\otimes Tr_{\mathcal{H}^3}$. Then (3) becomes

$S(\rho^{12}||\rho^1\otimes \rho^2)\leq S(\rho^{123}||\rho^1\otimes\rho^{23}).$

Therefore,

$S(\rho^{123}||\rho^1\otimes\rho^{23})- S(\rho^{12}||\rho^1\otimes \rho^2)=S(\rho^{12})+S(\rho^{23})-S(\rho^{123})-S(\rho^2)\geq 0,$

which is SSA. Thus, the monotonicity of quantum relative entropy (which follows from (1) implies SSA.

Owing to the Stinespring factorization theorem, equation (3) is valid not only for partial traces but also when $T$ is a quantum operation, i.e., a completely positive, trace preserving map. In this general case the inequality is called Monotonicity of quantum relative entropy.

Relationship among inequalities

All of the above important inequalities are equivalent to each other, and can also be proved directly. The following are equivalent:

• Monotonicity of quantum relative entropy (MONO);
• Monotonicity of quantum relative entropy under partial trace (MPT);
• Strong subadditivity (SSA);
• Joint convexity of quantum relative entropy (JC);

The following implications show the equivalence between these inequalities.

• MONO $\Rightarrow$ MPT: follows since the MPT is a particular case of MONO;

• MPT $\Rightarrow$ MONO: was showed by Lindblad,^[19] using a representation of stochastic maps as a partial trace over an auxiliary system;

• MPT $\Rightarrow$ SSA: follows by taking a particular choice of tri-partite states in MPT, described in the section above, "Monotonicity of quantum relative entropy";

• SSA $\Rightarrow$ MPT: by choosing $\rho_{123}$ to be block diagonal, one can show that SSA implies that the map

$\rho_{12}\mapsto S(\rho_1)-S(\rho_{12})$ is convex. In ^[3] it was observed that this convexity yields MPT;

• MPT $\Rightarrow$ JC: as it was mentioned above, by choosing $\rho_{12}$ (and similarly, $\sigma_{12}$) to be block diagonal matrix with blocks $\lambda_k\rho_k$ (and $\lambda_k\sigma_k$), the partial trace is a sum over blocks so that $\rho:=\rho_2=\sum_k\lambda_k\rho_k$, so from MPT one can obtain JC;

• JC $\Rightarrow$ SSA: using the 'purification process', Araki and Lieb,^[9][20] observed that one could obtain new useful inequalities from the known ones. By purifying $\rho_{123}$ to $\rho_{1234}$ it can be shown that SSA is equivalent to

$S(\rho_4)+S(\rho_2)\leq S(\rho_{12})+S(\rho_{14}).$

Moreover, if $\rho_{124}$ is pure, then $S(\rho_2)=S(\rho_{14})$ and $S(\rho_4)=S(\rho_{12})$, so the equality holds in the above inequality. Since the extreme points of the convex set of density matrices are pure states, SSA follows from JC;

See,^[20][21] for a discussion.

The case of equality

Equality in monotonicity of quantum relative entropy inequality

In,^[22][23] D. Petz showed that the only case of equality in the monotonicity relation is to have a proper "recovery" channel:

For all states $\rho$ and $\sigma$ on a Hilbert space $\mathcal{H}$ and all quantum operators $T: \mathcal{B}(\mathcal{H})\rightarrow \mathcal{B}(\mathcal{K})$,

$S(T\rho||T\sigma)= S(\rho||\sigma),$

if and only if there exists a quantum operator $\hat{T}$ such that

$\hat{T}T\sigma=\sigma,$ and $\hat{T}T\rho=\rho.$

Moreover, $\hat{T}$ can be given explicitly by the formula

$\hat{T}\omega=\sigma^{1/2}T^*\Bigl((T\sigma)^{-1/2}\omega(T\sigma)^{-1/2} \Bigr)\sigma^{1/2},$

where $T^*$ is the adjoint map of $T$.

D. Petz also gave another condition ^[22] when the equality holds in Monotonicity of quantum relative entropy: the first statement in Theorem below. Differentiating it at $t=0$ we have the second condition. Moreover, M.B. Ruskai gave another proof of the second statement.

For all states $\rho$ and $\sigma$ on $\mathcal{H}$ and all quantum operators $T: \mathcal{B}(\mathcal{H})\rightarrow \mathcal{B}(\mathcal{K})$,

$S(T\rho||T\sigma)= S(\rho||\sigma),$

if and only if the following equivalent conditions are satisfied:

• $T^*(T(\rho)^{it}T(\sigma)^{it})=\rho^{it}\sigma^{-it}$ for all real $t$.
• $\log\rho-\log\sigma=T^*\Bigl(\log T(\rho)-\log T(\sigma) \Bigr).$

where $T^*$ is the adjoint map of $T$.

Equality in strong subadditivity inequality

P. Hayden, R. Jozsa, D. Petz and A. Winter described the states for which the equality holds in SSA,.^[24]

A state $\rho^{ABC}$ on a Hilbert space $\mathcal{H}^A\otimes\mathcal{H}^B\otimes\mathcal{H}^C$ satisfies strong subadditivity with equality if and only if there is a decomposition of second system as

$\mathcal{H}^B=\bigoplus_j \mathcal{H}^{B^L_j}\otimes \mathcal{H}^{B^R_j}$

into a direct sum of tensor products, such that

$\rho^{ABC}=\bigoplus_j q_j\rho^{AB^L_j}\otimes\rho^{B^R_jC},$

with states $\rho^{AB^L_j}$ on $\mathcal{H}^A\otimes\mathcal{H}^{B^L_j}$ and $\rho^{B^R_jC}$ on $\mathcal{H}^{B^R_j}\otimes\mathcal{H}^C$, and a probability distribution $\{q_j\}$.

Operator extension of strong subadditivity

In his paper ^[25] I. Kim studied an operator extension of strong subadditivity, proving the following inequality:

For a tri-partite state (density matrix) $\rho^{123}$ on $\mathcal{H}^1\otimes \mathcal{H}^2\otimes\mathcal{H}^3$,

$Tr_{12}\Bigl(\rho^{123}(-\log(\rho^{12})-\log(\rho^{23})+\log(\rho^2)+\log(\rho^{123}))\Bigr) \geq 0.$

The proof of this inequality is based on Effros's theorem,^[26] for which particular functions and operators are chosen to derive the inequality above. M. B. Ruskai describes this work in details in ^[27] and discusses how to prove a large class of new matrix inequalities in the tri-partite and bi-partite cases by taking a partial trace over all but one of the spaces.

See also

• Von Neumann entropy
• Conditional quantum entropy
• Quantum mutual information
• Kullback–Leibler divergence

References

1. D. W. Robinson and D. Ruelle, Mean Entropy of States in Classical Statistical Mechanis, Communications in Mathematical Physics 5, 288 (1967)
2. O. Lanford III, D. W. Robinson, Jour. Mathematical Physics, 9, 1120 (1968)
3. E. H. Lieb, M. B. Ruskai, Proof of the Strong Subadditivity of Quantum Mechanichal Entropy, J. Math. Phys. 14, 1938–1941 (1973).
4. O. Fawzi, R. Renner. Quantum conditional mutual information and approximate Markov chains
5. M. Nielsen, I. Chuang Quantum Computation and Quantum Information, Cambr. U. Press, (2000)
6. M. Ohya, D. Petz, Quantum Entropy and Its Use, Springer (1993)
7. E. Carlen, Trace Inequalities and Quantum Entropy: An Introductory Course, Contemp. Math. 529 (2009).
8. H. Umegaki, Conditional Expectation in an Operator Algebra. IV. Entropy and Information, Kodai Math. Sem. Rep. 14, 59–85, (1962)
9. H. Araki, E. H. Lieb, Entropy Inequalities, Commun. Math. Phys. 18, 160–170 (1970).
10. Eric A. Carlen, Elliott H. Lieb, Bounds for Entanglement via an Extension of Strong Subadditivity of Entropy, Letters in Mathematical Physics, v.101, 1, 1-11, (2012)
11. E. H. Lieb, Convex Trace Function and Proof of Wigner–Yanase–Dyson Conjecture, Adv. Math. 11, 267–288 (1973).
12. H. Narnhofer, W.Thirring, From Relative Entropy to Entropy, Fizika 17, 258–262, (1985)
13. E. P. Wigner, M. M. Yanase, Information Content of Distributions, Proc. Nat. Acad. Sci. USA 49, 910–918 (1963).
14. E. P. Wigner, M. M. Yanase, On the Positive Semi-Definite Nature of a Certain Matrix Expression, Can. J. Math. 16, 397–406, (1964).
15. F. Hansen, The Wigner-Yanase Entropy is Not Subadditive, J. Stat. Phys. 126, 643–648 (2007).
16. A. Ulhmann, Endlich Dimensionale Dichtmatrizen, II, Wiss. Z. Karl-Marx-University Leipzig 22 Jg. H. 2., 139 (1973).
17. G. Lindblad, Expectations and Entropy Inequalities for Finite Quantum Systems, Commun. Math. Phys. 39, 111–119 (1974).
18. A. Ulhmann, Relative Entropy and the Wigner–Yanase–Dyson–Lieb Concavity in an Interpolation Theory, Comm. Math. Phys,54, 21–32, (1977).
19. G. Lindblad, Completely Positive Maps and Entropy Inequalities, Commun. Math. Phys. 40, 147–151 (1975).
20. E. H. Lieb, Some Convexity and Subadditivity Properties of Entropy, Bull. AMS 81, 1–13 (1975).
21. M. B. Ruskai, Inequalities for Quantum Entropy: A Review with Conditions for Equality, J. Math. Phys. 43, 4358–4375 (2002); erratum 46, 019901 (2005)
22. D. Petz, Sufficient Subalgebras and the Relative Entropy of States of a von Neumann Algebra, Commun. Math.Phys. 105, 123–131 (1986).
23. D. Petz, Sufficiency of Channels over von Neumann Algebras, Quart. J. Math. Oxford 35, 475–483 (1986).
24. P. Hayden, R. Jozsa, D. Petz, A. Winter, Structure of States which Satisfy Strong Subadditivity of Quantum Entropy with Equality, Comm. Math. Phys. 246, 359–374 (2003).
25. I. Kim, Operator Extension of Strong Subadditivity of Entropy, arXiv:1210.5190 (2012).
26. E. G. Effros. A Matrix Convexity Approach to Some Celebrated Quantum Inequalities. Proc. Natl. Acad. Sci. USA 106(4), 1006–1008 (2009).
27. M. B. Ruskai, Remarks on Kim’s Strong Subadditivity Matrix Inequality: Extensions and Equality Conditions, arXiv:1211.0049 (2012).