Professional Documents
Culture Documents
information thermodynamics with great success, see for with p(ξ, π) the joint probability. If the two vectors are
example25–32 . Thus the investigation of thermodynamics independent the joint entropy reduces to the sum of the
under feedback is rapidly rising interest in many fields individual entropies, in the other case we have:
beyond statistical physics.
In this paper we give a new insight to the problem of S(Ξ, Π) = S(Ξ) + S(Π) − kB I(Ξ, Π), (4)
a system under an external feedback using a special case
of LDT and the concept of typicality. The link between with I(Ξ, Π) the mutual information functional defined
systems under feedback and LDT is obtained using rate as:
distortion theory (RDT), a fundamental part of informa-
X p(ξ, π)
tion theory. The chain of relationships between LDT and I(Ξ, Π) = p(ξ, π) ln . (5)
RDT that we outline, is particularly pleasing as it makes p(ξ)p(π)
a direct connection between the information the feedback
The mutual information is a special case of the Kullback-
controller apparatus gathers about a system and the as-
Leibler divergence (KLd) defined as:
sociated entropy reduction, also leading to an explicit
construction of a thermodynamic potential for a system X
p(ξ)
under feedback. The possibility to construct potentials D(P kQ) = p(ξ) ln , (6)
q(ξ)
including the effect of information opens the perspective
of a complete new analysis of those systems more similar where P and Q are two PDFs. The KLd is a pseudo
to conventional equilibrium thermodynamic formalism. distance between PDFs and has an important role
For a similar approach see also33 . in LDT due to the Chernoff bound19 , in stochastic
The paper is organized as follows: in the first part a thermodynamics34 , within fluctuation theorems35 and in
brief summary of the main information theoretical con- entropy production within the Boltzmann equation36 . In
cepts, Shannon entropy, conditional entropy and mutual particular, we observe that the mutual information has
information is given. Then the concept of typicality is the form of a KLd between the joint PDF, p(ξ, π), and
introduced. In the next section we present the large devi- the independent PDF, p(ξ, π) = p(ξ)p(π) ≡ q(ξ, π).
ation theory and the code large distortion problem. The Finally, mutual information and entropy of two ran-
final part is devoted to introducing rate distortion theory dom vectors are connected by the conditional entropy:
and the final link to thermodynamics under an external
feedback. X
S(Ξ|Π) = S(Ξ) − kB I(Ξ, Π) = p(ξ, π) ln p(ξ|π). (7)
II. A BRIEF EXCURSUS IN INFORMATION Conditional entropy represents the entropy (uncertainty)
THEORY left in Ξ after the conditioning to Π.
These five are the most relevant functionals in infor-
The central quantity of information theory is the Shan- mation theory as practically every important theorem is
non entropy defined as related to one or another in some form.
X
S(P ) = −kB pi ln pi , (2)
i III. TYPICAL SET IN THE PHASE SPACE
th
where pi represents the probability of the i event and
P the PDF. Shannon entropy is usually expressed in bits The Shannon entropy has a nice geometrical inter-
and adimensional, we have chosen here the convention to pretation in the typical set theorem consequence of the
include the Boltzmann constant and express the entropy asymptotic equipartition principle37 . The theorem states
using natural logarithm. However, this is totally imma- that given a PDF p(ξ), where ξ is the vector of degrees
terial for the present discussion. Shannon entropy plays of freedom of the problem, with entropy S(Ξ) then the
a central role in thermodynamics, even for a large class ”typical set” within the phase space has a volume equal
of systems under non equilibrium conditions16 . to:
More generally if we have a PDF p(ξ) which depends S(Ξ)
on a set of degrees of freedom, ξ, we can define the proper Ωtyp ∼ e kB
, (8)
discrete (or continuous) Shannon entropy. The degrees
of freedom depend on the problem at hand, they could where the meaning of ”typical” means that the probabil-
be the set of positions and momenta of a collection of ity for the system to be found in a microstate within the
particle in gas phase for example. From the Shannon typical set converges to 1 in the thermodynamic limit,
entropy it is possible to promptly derive other four con- namely,
nected quantities. The first is the joint (Shannon) en-
tropy which is just the entropy of two random vectors ξ p(ξ ∈ Ωtyp ) → 1. (9)
and π:
X In other words the concept of typicality states that only
S(Ξ, Π) = −kB p(ξ, π) ln p(ξ, π), (3) a portion of the entire phase space is really relevant to
3
with
If it happens that for all microstates within the typical
set, A(ξ ∈ Ωtyp ) = A∗ (constant), then A is a state
Z
N αa
variable of the problem and hAi = A∗ . e = p(ξ)eαN a(ξ) dξ, (14)
Several generalizations of the typical set concept ex-
ist for example for a joint PDF in the joint typical set with α a real number. The GET states that if the SCGF
theorem37 . The concept of typicality is extremely im- exists and is differentiable everywhere in α, then the sys-
portant also in thermodynamics. In the microcanonical tem fluctuations follow an exponential law.
ensemble where the PDF over the phase space is a uni- We notice that the SCGF is very similar to a partition
form PDF, the entropy reduces to the integration of the function, but scaled with respect to N and also where
density of accessible microstates. every exponential term is weighted by the probability per
However, as recent studies on stochastic thermody- microstate, p(ξ). If we make a simple variable change
namics have shown15,34 , it is possible to extend many α = −β, the new parameter β plays the same role as
thermodynamic results also for small systems, i.e., sys- the inverse temperature, β = 1/(kB T ) and the SCGF
tems where significant fluctuations of the state variables can be rewritten as −φ(β) = λ(α). The GET does not
are not only possible, but also probable. Such type of only provide a condition for existence, but also gives an
systems are also those for which a practical implementa- operative way to evaluate the exponent K(a). In fact
tion of a feedback controller is more feasible due to their it demonstrates that K(a) is related to the Legendre-
limited number of degrees of freedom. Fenchel transform of the SCGF:
SCGF), but adding a constant: A dual theorem of GET is the Varadhan theorem48
which allows to invert the Legendre-Fenchel transform.
1 1
J(a) = − min [βa − φ(β)] + ln Λ = K(a) + ln Λ. This states that if
β≥0 N N
(16) K(a) = − min [βa − φ(β)], (22)
The latter constant has the form of the entropy of a uni- β≥0
form PDF U (ξ) = 1/Λ, with Λ a particular volume within
the phase space (see44 ), which depends on the prior PDF is valid, then also the following relation holds:
p(ξ). In the case of a microcanonical ensemble it reduces
to all the microstates with the same energy Ē. The func- φ(β) = min[βa + K(a)]. (23)
a
tion φ(β) is related to the free energy potential. Specifi-
cally, we have that the free energy per degrees of freedom The LD law in microcanonical or canonical ensemble ex-
(f = F/N ) is equal to: plains, through the Varadhan theorem, why the minima
of the free energy potentials are associated to the state
φ(β) variable values for the equilibrium state.
f= , (17)
β
thus the Legendre-Fenchel transform of the SCGF is
linked to the entropy of the system per degree of free- V. RATE DISTORTION THEORY AND LARGE
DEVIATION THEORY
dom, s(a), by
s(a) 1 In this paragraph we make explicit the connection be-
J(a) = − + ln Λ. (18)
kB N tween LDT and the information theory quantities pre-
sented in the previous sections and consider a system
If we consider the constant term as the entropy associ-
under feedback. In order to do so we need to introduce
ated to a uniform PDF we have that the exponent J(a)
the rate distortion function (RDF), a central functional
can be treated as follows:
in rate distortion theory (RDT). RDT copes with a fun-
−s(a) 1 damental problem in communication, namely estimating
J(a) = + ln Λ
kB N the minimal information content that any message sent
over a communication channel must contain such that
1 X X
= p(ξ) ln p(ξ) + p(ξ) ln Λ the receiver can still have a good reconstruction of the
N original signal. The error source can be either due to
ξ ξ
1 distorting noise, a finite channel capacity or because of
= D(Ξ//U ), (19) a lossy compression operated by the sender. In mathe-
N matical form, if we define a distance, d(π, ξ), between the
where U = 1/Λ is the uniform PDF over the phase message sent, ξ, and the message received, π (eventually
space
P volume Λ, D is the KLd and N s(a) = Stot = after decompression, noise deconvolution, etc...), the tar-
− p(ξ) ln p(ξ) the total entropy. Thus the probability get of RDT is to find under which conditions the average
q(a 6= a∗ ) goes like the following: distance can be kept lower than a given threshold, Γ37 .
Formally we ask,
q(a 6= a∗ ) ≈ e−D(Ξ//U) , (20)
recovering the Chernoff bound for large fluctuations hd(ξ, π)i ≤ Γ, (24)
within the typical set formalism37 . The LDT provides a
clear understanding why entropy maximization is at the where the average is computed over the joint PDF p(ξ, π).
essence of equilibrium thermodynamics. Similar results The distance, d, can be any functional with the properties
are obtained for a canonical ensemble18 . For a general of a distance (symmetry, positive definite, d = 0 iif ξ =
discussion about the GET and the LDT applied to ther- π, and must satisfy the Schwartz inequality). The most
modynamics we refer to39 . important theorem of RDT states that, given d and Γ,
Notably, LDT has been applied to non equilibrium there exists a function, R(Γ) (the rate distortion func-
systems38 substituting the PDF for a microstate with tion), representing the minimum information required in
the PDF of entire time dependent trajectories to take order to send messages with an average distortion not
into account the time evolution of the system. Even more greater than Γ. The function R(Γ) has some remarkable
important, LDT can be used to derive fluctuation theo- properties: it is convex in the argument Γ, it converges to
rems. In fact if the exponent has the symmetry relation, the entropy of the source (for a discrete PDF) for Γ = 0,
e.g. K(−a) − K(a) = γa (γ is a positive real number) while for Γ ≥ Γ∗ it is zero. In practice Γ∗ is the limit
for a certain thermodynamic quantity, A = N a , we im- distortion value after which the information content is
mediately get the fluctuation theorem18,34,45–47 : completely lost and the receiver has equal chance by just
guessing at random the most likely message, based on the
q(a) joint PDF37 . For an example of a RDF for a Gaussian
≈ eN (K(−a)−K(a)) = eN γa . (21)
q(−a) PDF see Fig. 2.
5
of the feedback is to compress the volume of the typical system accordingly in a time τ much smaller than any re-
set with an exponent at best equal to the mutual infor- laxation time of the system. Moreover, we assume a cer-
mation: tain level of ideality in the feedback, i) during the mea-
surement no perturbation of the system occurs, ii) the
Ωtyp (ξ|π) ≥ Ωtyp (ξ)e−I(Ξ,Π) . (35) feedback uses the entire information during the manipu-
lation achieving the maximal efficiency. Once the system
The fact that the mutual information is always non has been manipulated by the feedback it is allowed to
negative ensures that the effect of the feedback is always relax and finally it is connected again to the external
to compress the original typical set. bath until thermalize with it. This simple cycle assures
If we want to connect the mutual information to the that every time the feedback performs its new measure-
effective measurement operated by the feedback appa- ment/manipulation the system is at thermal equilibrium
ratus (FA), then we can define a distance functional, d, with temperature T0 .
and an average distortion Γ between state and estimation With these approximations we can decouple the initial
and use the rate distortion function to find the minimal state before the feedback operation from the feedback
mutual information required to have a certain average action. A completely different and more complex sce-
distortion Γ. The final result is the lower boundary, nario occurs in the case when the system is not allowed
to thermalize or the feedback acts continuously in such a
hΩtyp (ξ|π = πk )i ≥ Ωtyp (ξ)e−R(Γ) . (36) way that the system state depends on previous feedback
history.
The appealing of equation (36) is that it gives a direct We finally assume that the feedback can obtain an es-
connection between the effect of feedback information timate, π, of the real state, ξ, of the system such that
and the effective operative measurement performed by hd(ξ, π)i ≤ Γ for a well-defined distance functional.
the FA. We have shown in section 6 that the feedback action
A final note regarding the effect of the FA. The re- entails to an entropy reduction after measurement by a
duction in the volume of the typical set remains only factor I(ξ, π) = R(Γ). Thus we can finally define the
”virtual” until the FA does not operate and manipulate maximum increase in free energy due to the presence of
the system. We can thus imagine this shrinking of the the feedback:
typical set linked to the R(Γ) as the best average reduc-
tion in uncertainty about the state of the system from ∆F = F − Feq
the FA side once the measurements have been made. It = U − T0 (S − kB R(Γ)) − U + T0 S = kB T0 R(Γ) = hW i.
is anyway clear that the entropy and the physics of the (37)
system are left unchanged until the FA does not directly
operate. with U the internal energy. It recovers the result found
by Sagawa with eq. 1, as the maximum work for a cyclic
transformation, but now with a direct connection be-
VII. TOWARDS A THERMODYNAMIC
tween the type and accuracy of the measurement, d(ξ, π)
POTENTIAL INCLUDING INFORMATION and Γ, and the average extracted work hW i.
The R(Γ) exponent of equation (36), following a LD
law, can be expanded in the Legendre-Fenchel transform
A thermodynamic potential is nothing else than a as in equation (15), leading to a joint probability between
quantity that when minimized/maximized allows to find the microstate and the guess equal to equation (29). Us-
the equilibrium or stationary states of a certain thermo- ing the Varadhan theorem we can invert the Legendre-
dynamic system, subject to given constrains. We have Fenchel transform for the RDF and obtain something
already discussed in section 4 and 5 how the LDT pro- equivalent to a thermodynamic potential for the infor-
vides an elegant way to derive the maximal entropy and mation:
minimal energy principles for the microcanonical and
canonical ensembles. In particular, we saw how the free
X
hln(Zπ (λ)i = p(ξ) ln(Zπ (λ)) = min [λΓ + R(Γ)] .
energy potential comes as part of the Legendre-Fenchel Γ
transform of the large deviation exponent thanks to the (38)
Gartner-Ellis theorem. In this section we put together The structure of this equation is similar to the one of the
this result and the results of section 6 in order to con- free energy in standard thermodynamics with the rela-
struct an explicit thermodynamic potential for a system tion between free energy and partition function, where
under feedback control. now the Chernoff coefficient λ plays the role of inverse
First of all we define a simplified feedback apparatus temperature:
(SFA). We assume a system in thermal equilibrium with 1
an external bath at temperature T0 and we assume that φI (λ) ≡ − hln(Zπ (λ)i . (39)
λ
the system can be coupled to the bath or disconnected
from that and coupled to the SFA. The SFA is a sys- What is the meaning of this functional? We can un-
tem that can make a measurement and manipulate the derstand its role by calculating the distance, using the
7
X p(ξ, π)
D(P//P̃ ) = p(ξ, π) ln
p̃(ξ, π)
ξπ
X p(ξ, π) FIG. 3. (color online) Scheme of a channel between two joint
= p(ξ, π) log + λ∗ hd(ξ, π)i + hln Zπ (λ∗ )i
p(ξ)q(π) gaussian random variables.
ξπ
= I(Ξ, Π) + λ∗ hd(ξ, π)i − λ∗ φI (λ∗ ). (40)
The latter equation can be rewritten as: Following37 the feedback has a PDF equals to:
2
λ∗ φI (λ∗ ) = λ∗ hd(ξ, π)i + I(Ξ, Π) − D(P//P̃ ). (41) 1 − π2
p(π) = N (0, σξ2 − Γ) = q e 2(σξ −Γ) . (43)
2π(σξ2 − Γ)
Considering that the three terms in the rhs are all posi-
tive it is easy to demonstrate that the maximum of the
potential φI (λ∗ ) is obtained for D(P//P̃ ) = 0, that is In the latter we have assumed that 0 ≤ Γ ≤ σξ2 . With the
when the joint probability is the ideal one for which the previous defined distance functional the average distance
RDF is achieved. Γ is equal to the mean square error h(ξ − π)2 i.
For this simple model the form of the rate distortion
function is well known:
VIII. A SIMPLE PHYSICAL MODEL !
1 σξ2
R(Γ) = ln , (44)
2 Γ
We apply the formalism to a simple model: a set of sin-
gle particles in a box in gas phase. We assume that every
particle is under a feedback and that it is in thermal equi- with R(Γ) = 0 for Γ ≥ σξ2 .
librium with a bath at temperature T . The Hamiltonian It is a simple matter of calculation to demonstrate that
of the system is given by E = Ap2 , where A = 1/2m, indeed eq. 27, using p(ξ), p(π) and d(ξ, π) = (ξ − π)2 ,
being m the particle mass. The particle state, ξ = (r, p), gives back the correct rate distortion function. The rela-
is characterized by an homogeneous spatial distribution tion between λ and Γ is very simple:
for the position r within the box, and a normal distribu- 1
tion of the momentum, p. Therefore, neglecting position, λ= . (45)
2Γ
we identify the particle state just with the momentum,
x = p. The equilibrium distribution is a normal distri- We can insert the PDF and the distance functional in
bution with 0 mean value and a variance σξ2 = kB T /2A. eq. 30 in order to get the optimal joint distribution for
The model includes a feedback that can probe the mo- such marginal PDFs. The solution is a joint Gaussian
mentum of the particle. The feedback uses a distance distribution:
d(ξ, π) = (ξ − π)2 . This measurement is affected by er-
ror, so the measurement π is a random variable, statis- p̃(ξ, π) =
" #!
tically correlated with the dynamical state of the parti- 1 1 ξ2 2ρξπ π2
cle. The model assumes that between every probe and exp − − + 2
2(1 − ρ2 ) σξ2
p
2πσξ σπ 1 − ρ2 σξ σπ σπ
manipulation the system is allowed to relax to thermal
equilibrium, that means that at every measurement the (46)
system is found in the same equilibrium state. Finally, q
we also assume that π is distributed like a normal ran- with correlation coefficient equal to ρ = 1 − Γ/σξ2 . Us-
dom variable. The situation can be formalized like in37 ing the result in eq. 45 we obtain the relation between λ
by assuming that the feedback and the source are con- and the correlation coefficient:
nected by a channel with Gaussian noise (see Fig. 3).
1
λ= . (47)
The Gaussian noise has distribution: 2σξ2 (1 − ρ2 )
Γ has an extensive characteristic being function of the accuracy and the maximum work that can be extracted in
number of degree of freedom (N ) of the system. On the terms of entropy reduction. Clearly in this discussion we
contrary λ is the intensive counterpart being connected have not considered the effect of the manipulation from
to the correlation between state and estimation. the feedback and its cost in terms of overall efficiency,
In this particular case we can also calculate the optimal thus all our results must be considered as boundary lim-
work that such a feedback apparatus can recover from its.
the system as a function of the average distance Γ, using The main result of the paper is the possibility, us-
eq.37: ing rate distortion theory, of developing the equivalent
of thermodynamic potentials even in the case of systems
T kB
hW i = T kB R(Γ) = − ln(1 − ρ2 ) = T kB I(Ξ, Π). under an external feedback.
2
(48) This perspective to the problem is particular appealing
The latter shows that the maximum work extracted is not only because it establishes a nice relation between the
proportional as should be to the mutual information be- measurement and the respective entropy and information
tween two Gaussian random variables. functional, but also because it can be generalized for a
large class of problems where large deviation applies.
IX. CONCLUSIONS
X. ACKNOWLEDGEMENTS
In this work we have analyzed a system under feedback
using typicality and large deviation theory. In particular
this connection assures to make a simple and natural link We acknowledge Prof. Merhav for the useful com-
between the way the measurement is performed and its ments.
1 16
H. S. Leff, A. F. Rex, ”Maxwell Demon: Entropy, Informa- J. M. R. Parrondo, J. M. Horowitz and T. Sagawa, Nature
tion, Computing, Princeton University Press”, Princeton, Phys., vol. 11, 131 (2015).
17
NJ, (1990). A. Gagliardi and A. Pecchia, arXiv:1503.02824v1 (2015).
2 18
H. S. Leff, A. F. Rex, ”Maxwell Demon 2: Entropy, Clas- H. Touchette, Phys. Rep., vol. 478, 1-69 (2009).
19
sical and Quantum Information, Computing”, Institute of N. Merhav, IEEE Trans. Inform. Theory, vol. 54, no. 8,
Physics, Bristol, (2003). pp. 3710-3721 (2008).
3 20
L. Szilard, On the Decrease of Entropy in a Thermody- A. Kis Andras and A. Zettl, Philos. Trans. R. Soc. A, 366,
namic System by the Intervention of Intelligent Beings, Z. 1591 (2008).
21
Phys. 53:840 (1929): English translation reprinted Behav- J. R. Gomez-Solano, L. Bellon, A. Petrosyan and S. Cilib-
ioral Science 9:301 (1964). erto, Europhys. Lett., 89, 60003 (2010).
4 22
H. Touchette and S. Lloyd,Phys. Rev. E, 84, 1156 (2000). L. Bellon, L. Buisson, S. Ciliberto and F. Vittoz, Rev. Sci.
5
A. E. Allahverdyan, D. Janzing, and G. Mahler,J. Stat. Instrum., 73, 3286 (2002).
23
Mech., P09011 (2009). A. Berut, A. Arakelyan, A. Petrosyan, S. Ciliberto, R. Dil-
6
J. Horowitz, T. Sagawa, and J. M. R. Parrondo,Phys. Rev. lenschneider and E Lutz, Nature, 187, 483 (2012).
24
Lett., 111, 010602 (2013). J. V. Koski, V. F. Maisi, T. Sagawa and J. P. Pekola, Phys.
7
J. M. Horowitz and M. Esposito,Phys. Rev. X, 4, 031015 Rev. Lett., 113, 030601 (2014).
25
(2014). A. C. Barato, D. Hartich and U. Seifert, Phys. Rev.E, 87,
8
J. M. Horowitz and H. Sandberg,New J. Phys., 16, 125007 042104 (2013).
26
(2014). S. Ito, T. Sagawa, Nat. Comm., 6, 7498 (2015).
9 27
N. Shiraishi, S. Ito, K. Kawaguchi, and T. Sagawa,New J. A. H. Lang, C. K. Fisher, T. Mora and P. Mehta, Phys.
Phys., 17, 045012 (2015). Rev. Lett., 113, 148103 (2014).
10 28
D. Hartich, A. C. Barato, and U. Seifert,J. Stat. Mech., P. Sartori, L. Granger, C. F. Lee and J. M. Horowitz, PLoS
P02016 (2014). Comput. Biol., 10, 1003974 (2014).
11 29
S. Toyabe, T. Sagawa, M. Ueda, E. Muneyuki and M. R. G. Endres and N. S. Wingreen, Phys. Rev. Lett., 103,
Sano,Nature Physics, 6, 988 (2010). 158101 (2009).
12 30
T. Sagawa and M. Ueda, Phys. Rev. Lett., 100, 080403 H. Qian and T. C. Reluga, Phys. Rev. Lett., 94, 028101
(2008). (2005).
13 31
T. Sagawa and M. Ueda, Phys. Rev. Lett., vol. 104, 090602 P. Mehta and D. Schwab, Proc. Natl Acad. Sci. USA, 109,
(2010). 17978 (2012).
14 32
T. Sagawa and M. Ueda, Chap. 6, ”Nonequilibrium Sta- Y. Tu,Proc. Natl Acad. Sci. USA, 105, 11737 (2008).
33
tistical Physics of Small systems”, (Wiley-VCH), (2013). N. Merhav, J. Stat. Phys., P01029, doi: 10.1088/1742-
15
S. Hilbert, P. Haenggi and J. Dunkel, Phys. Rev. E, vol. 5468/2011/01/P01029 (2011).
34
90, 062116 (2014). U. Seifert, Rep. Prog. Phys., 75, 126001 (2012).
9
35
C. Jarzynski, Non-equilibrium equality for free energy dif- Berlin, pp. 1-113 (1973).
44
ferences, Phys. Rev. Lett., vol. 78, 2690 (1997). N. Merhav, IEEE IST (2008) Toronto, 499 (2008).
36 45
F. Rezakhanlou and C. Villani, ”Entropy Methods for the D. J. Evans and D. J. Searles, Phys. Rev. E, 50, 1645
Boltzmann Equation”, Springer-Verlag Berlin Heidelberg (1994).
46
(2008). G. Gallavotti and E. G. D. Cohen, Phys. Rev. Lett., 74,
37
T. M. Cover and J. A. Thomas, ”Elements of information 2694 (1995).
47
theory”, Wyley (2006). J. L. Lebowitz and H. Spohn, J. Stat. Phys., 95, 333 (1999).
38 48
H. Touchette, R. J. Harris, Chap. 11, ”Nonequilibrium Sta- S. R. S. Varadhan, Comm. Pure App. Math., 19, 261
tistical Physics of Small systems”, (Wiley-VCH), (2013). (1966).
39 49
R. S. Ellis, ”Entropy, Large Deviations and Statistical Me- N. Merhav, Statistical Physics and Information theory,
chanics”, Springer, New York (1985). Foundation and Trends in Communication and Informa-
40
R. S. Ellis, Physica D, 133, 106-136 (1999). tion Theory, vol. 6, 1-212 (2009).
41 50
Y. Oono, Progr. Theoret. Phys. Suppl. 99, 165-205 (1989). T. Berger, ”Rate distortion theory: a mathematical ba-
42
R. S. Ellis, Scand. Actuar. J., 1, 97-142 (1995). sis for data compression”, PrenticeHall, Inc., Engelwood
43
O. E. Lanford, ”Entropy and equilibrium states in classi- Cliffs, NJ, (1971).
51
cal statistical mechanics”, in: A. Lenard (Ed.), Statistical R. M. Gray, ”Source Coding Theory”, Kluwer Academic
Mechanics and Mathematical Problems, vol. 20, Springer, Publishers, (1990).