Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.
-
date post
20-Dec-2015 -
Category
Documents
-
view
241 -
download
4
Transcript of Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.
![Page 1: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/1.jpg)
Shannon’s theory
Ref. Cryptography: theory and practice
Douglas R. Stinson
![Page 2: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/2.jpg)
Shannon’s theory
1949, “Communication theory of Secrecy Systems” in Bell Systems Tech. Journal.
Two issues: What is the concept of perfect secrecy? Does
there any cryptosystem provide perfect secrecy?
It is possible when a key is used for only one encryption
How to evaluate a cryptosystem when many plaintexts are encrypted using the same key?
![Page 3: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/3.jpg)
Outline
Introduction One-time pad
Elementary probability theory Perfect secrecy Entropy Spurious keys and unicity distance
![Page 4: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/4.jpg)
Categories of cryptosystem (1)
Computational security: The best algorithm for breaking a cryptosyste
m requires at least N operations, where N is a very large number
No known practical cryptosystem can be proved to be secure under this definition
Study w.r.t certain types of attacks (ex. exhaustive key search) does not guarantee security against other type of attack
![Page 5: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/5.jpg)
Categories of cryptosystem (2)
Provable security Reduce the security of the cryptosystem to
some well-studied problems that is thought to be difficult
Ex. RSA integer factoring problem Unconditional security
A cryptosystem cannot be broken, even with infinite computational resources
![Page 6: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/6.jpg)
One-Time Pad
Unconditional security !!! Described by Gilbert Vernam in 1917 Use a random key that was truly as
long as the message, no repetitionsnKCP )( 2 ),,( 1 nxxx ),,( 1 nKKK
2 mod ),,()( 11 nnK KxKxxe
2 mod ),,()( 11 nnK KyKyyd
For ciphertext ),,( 1 nyyy
![Page 7: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/7.jpg)
Example: one-time pad
Given ciphertext with Vigenère Cipher: ANKYODKYUREPFJBYOJDSPLREYIUNOFDOIUERFPLUYTS
Decrypt by hacker 1:
Ciphertext: ANKYODKYUREPFJBYOJDSPLREYIUNOFDOIUERFPLUYTSKey: pxlmvmsydofuyrvzwc tnlebnecvgdupahfzzlmnyihPlaintext: mr mustard with the candlestick in the hall
Decrypt by hacker 2:
Ciphertext: ANKYODKYUREPFJBYOJDSPLREYIUNOFDOIUERFPLUYTSKey: pftgpmiydgaxgoufhklllmhsqdqogtewbqfgyovuhwtPlaintext: miss scarlet with the knife in the library
Which one?
![Page 8: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/8.jpg)
a b c d e f g h i j k l m n o p q r s t u v w x y z ?A B C D E F G H I J K L M N O P Q R S T U V W X Y Z ?B C D E F G H I J K L M N O P Q R S T U V W X Y Z ? AC D E F G H I J K L M N O P Q R S T U V W X Y Z ? A BD E F G H I J K L M N O P Q R S T U V W X Y Z ? A B CE F G H I J K L M N O P Q R S T U V W X Y Z ? A B C DF G H I J K L M N O P Q R S T U V W X Y Z ? A B C D EG H I J K L M N O P Q R S T U V W X Y Z ? A B C D E FH I J K L M N O P Q R S T U V W X Y Z ? A B C D E F GI J K L M N O P Q R S T U V W X Y Z ? A B C D E F G HJ K L M N O P Q R S T U V W X Y Z ? A B C D E F G H IK L M N O P Q R S T U V W X Y Z ? A B C D E F G H I JL M N O P Q R S T U V W X Y Z ? A B C D E F G H I J KM N O P Q R S T U V W X Y Z ? A B C D E F G H I J K LN O P Q R S T U V W X Y Z ? A B C D E F G H I J K L MO P Q R S T U V W X Y Z ? A B C D E F G H I J K L M NP Q R S T U V W X Y Z ? A B C D E F G H I J K L M N OQ R S T U V W X Y Z ? A B C D E F G H I J K L M N O PR S T U V W X Y Z ? A B C D E F G H I J K L M N O P QS T U V W X Y Z ? A B C D E F G H I J K L M N O P Q RT U V W X Y Z ? A B C D E F G H I J K L M N O P Q R SU V W X Y Z ? A B C D E F G H I J K L M N O P Q R S TV W X Y Z ? A B C D E F G H I J K L M N O P Q R S T UW X Y Z ? A B C D E F G H I J K L M N O P Q R S T U VX Y Z ? A B C D E F G H I J K L M N O P Q R S T U V WY Z ? A B C D E F G H I J K L M N O P Q R S T U V W XZ ? A B C D E F G H I J K L M N O P Q R S T U V W X Y? A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
abcdefghijklmnopqrstuvwxyz?
![Page 9: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/9.jpg)
Problem with one-time pad
Truly random key with arbitrary length? Distribution and protection of long keys
The key has the same length as the plaintext!
One-time pad was thought to be unbreakable, but there was no mathematical proof until Shannon developed the concept of perfect secrecy 30 years later.
![Page 10: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/10.jpg)
Preview of perfect secrecy (1)
When we discuss the security of a cryptosystem, we should specify the type of attack that is being considered Ciphertext-only attack
Unconditional security assumes infinite computational time Theory of computational complexity × Probability theory ˇ
![Page 11: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/11.jpg)
Preview of perfect secrecy (2)
Definition: A cryptosystem has perfect secrecy if Pr[x|y] = Pr[x] for all xP, yC
Idea: Oscar can obtain no information about the plaintext by observing the ciphertext
Alice Bob
Oscar
x y
![Page 12: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/12.jpg)
Outline
Introduction One-time pad
Elementary probability theory Perfect secrecy Entropy Spurious keys and unicity distance
![Page 13: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/13.jpg)
Discrete random variable (1)
Def: A discrete random variable, say X, consists of a finite set X and a probability distribution defined on X.
The probability that the random variable X takes on the value x is denoted Pr[X=x] or Pr[x]
0≤Pr[x] for all xX, 1]Pr[ Xx
x
![Page 14: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/14.jpg)
Discrete random variable (2)
Ex. Consider a coin toss to be a random variable defined on {head, tails} , the associated probabilities Pr[head]=Pr[tail]=1/2
Ex. Throw a pair of dice. It is modeled by Z={(1,1), (1,2), …, (2,1), (2,2), …, (6,6)}
where Pr[(i,j)]=1/36 for all i, j. sum=4 corresponds to {(1,3), (2,2), (3,1)} with
probability 3/36
![Page 15: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/15.jpg)
Joint and conditional probability
X and Y are random variables defined on finite sets X and Y, respectively.
Def: the joint probability Pr[x, y] is the probability that X=x and Y=y
Def: the conditional probability Pr[x|y] is the probability that X=x given Y=y
Pr[x, y] =Pr[x|y]Pr[y]= Pr[y|x]Pr[x]
![Page 16: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/16.jpg)
Bayes’ theorem If Pr[y] > 0, then
Ex. Let X denote the sum of two dice. Y is a random variable on {D, N}, Y=D i
f the two dice are the same. (double)
]Pr[
]|Pr[]Pr[]|Pr[
y
xyxyx
3
1
36/3
)6/1)(6/1(
]4Pr[
]Pr[]|4Pr[]4|Pr[
DDD
![Page 17: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/17.jpg)
Outline
Introduction One-time pad
Elementary probability theory Perfect secrecy Entropy Spurious keys and unicity distance
![Page 18: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/18.jpg)
Definitions
Assume a cryptosystem (P,C,K,E,D) is specified, and a key is used for one encryption
Plaintext is denoted by random variable x Key is denoted by random variable K Ciphertext is denoted by random variable y
Plaintext Ciphertext
x y
K
![Page 19: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/19.jpg)
Perfect secrecy
Definition: A cryptosystem has perfect secrecy if Pr[x|y] = Pr[x] for all xP, yC
Idea: Oscar can obtain no information about the plaintext by observing the ciphertext
Alice Bob
Oscar
x y
![Page 20: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/20.jpg)
Relations among x, K, y
Ciphertext is a function of x and K
y is the ciphertext, given that x is the plaintext
)}(:{
)](Pr[]Pr[]Pr[KCyK
K ydKy xKy
)}(:{
]Pr[]|Pr[ydxK K
Kxy Kxy
![Page 21: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/21.jpg)
Relations among x, K, y
x is the plaintext, given that y is the ciphertext
]|Pr[ yx yx]Pr[
]|Pr[]Pr[
y
xyx
)}(:{
)}(:{
)](Pr[]Pr[
]Pr[]Pr[
KCyKK
ydxK
ydK
KxK
xK
Kx
![Page 22: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/22.jpg)
Ex. Shift cipher has perfect secrecy (1)
Shift cipher: P=C=K=Z26 , encryption is defined as
Ciphertext:
26 mod )()( KxxeK
26
)](Pr[]Pr[]Pr[ZK
K ydKy xKy
26
]Pr[261
ZK
Kyx
261
261
26
]Pr[ ZK
Kyx
![Page 23: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/23.jpg)
Ex. Shift cipher has perfect secrecy (2)
Pr[y|x]
Apply Bayes’ theorem
261]26 mod )(Pr[ xyK
]Pr[
]|Pr[]Pr[]|Pr[
y
xyxyx
]Pr[]Pr[
261
261
xx
Perfect secrecy
![Page 24: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/24.jpg)
Perfect secrecy when |K|=|C|=|P|
(P,C,K,E,D) is a cryptosystem where |K|=|C|=|P|.
The cryptosystem provides perfect secrecy iff every keys is used with equal probability 1/|K| For every xP, yC, there is a unique key K suc
h that
Ex. One-time pad in Z2
yxeK )(
![Page 25: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/25.jpg)
Outline
Introduction One-time pad
Elementary probability theory Perfect secrecy Entropy Spurious keys and unicity distanceHow about the security when many plaintexts are encrypted using one key?
![Page 26: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/26.jpg)
Preview (1)
We want to know: the average amount of ciphertext required
for an opponent to be able to uniquely compute the key, given enough computing time
Plaintext Ciphertextxn yn
K
![Page 27: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/27.jpg)
Preview (2)
We want to know: How much information about the key is rev
ealed by the ciphertext = conditional entropy H(K|Cn)
We need the tools of entropy
![Page 28: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/28.jpg)
Entropy (1)
Suppose we have a discrete random variable X What is the information gained by the outcome
of an experiment? Ex. Let X represent the toss of a coin, Pr[he
ad]=Pr[tail]=1/2 For a coin toss, we could encode head by 1,
and tail by 0 => i.e. 1 bit of information
![Page 29: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/29.jpg)
Entropy (2) Ex. Random variable X with Pr[x1]=1/2,
Pr[x2]=1/4, Pr[x3]=1/4 The most efficient encoding is to
encode x1 as 0, x2 as 10, x3 as 11. Notice: probability 2-n => n bits p => -log2 p The average number of bits to encode
X2
32
4
12
4
11
2
1
![Page 30: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/30.jpg)
Entropy: definition
Suppose X is a discrete random variable which takes on values from a finite set X. Then, the entropy of the random variable X is defined as
Xx
xxH ]Pr[log]Pr[)( 2X
![Page 31: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/31.jpg)
Entropy : example
Let P={a, b}, Pr[a]=1/4, Pr[b]=3/4. K={K1, K2, K3}, Pr[K1]=1/2, Pr[K2]=Pr[K3]= 1/4. encryption matrix: a b
K1 1 2
K2 2 3
K3 3 4
H(P)= 81.04
3log
4
3
4
1log
4
122
H(K)=1.5, H(C)=1.85
![Page 32: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/32.jpg)
Conditional entropy
Known any fixed value y on Y, information about random variable X
Conditional entropy: the average amount of information about X that is revealed by Y
Theorem: H(X,Y)=H(Y)+H(X|Y)
x
yxyxyH ]|Pr[log]|Pr[)|( 2X
y x
yxyxyH ]|Pr[log]|Pr[]Pr[)|( 2YX
![Page 33: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/33.jpg)
Theorem (1)
Let (P,C,K,E,D) be a cryptosystem, then H(K|C) = H(K) + H(P) – H(C) Proof:
H(K,P,C) = H(C|K,P) + H(K,P)
Since key and plaintext uniquely determine the ciphertext
H(C|K,P) = 0
H(K,P,C) = H(K,P) = H(K) + H(P)
Key and plaintext are independent
![Page 34: Shannon ’ s theory Ref. Cryptography: theory and practice Douglas R. Stinson.](https://reader035.fdocuments.net/reader035/viewer/2022081504/56649d485503460f94a237ad/html5/thumbnails/34.jpg)
Theorem (2) We have Similarly,
Now,
H(K,P,C) = H(K,C) = H(K) + H(C)
H(K,P,C) = H(K,P) = H(K) + H(P)
H(K|C)= H(K,C)-H(C)
= H(K,P,C)-H(C)
= H(K)+H(P)-H(C)