The Cipher Code Method- Thinking after the Freshman Discussion Lectures on Image Processing-CSDN博客

本文链接：https://blog.csdn.net/qq_16045611/article/details/84315898

电子科技大学格拉斯哥学院 2017级王奉彦

Wu Fengyi , Wang Fengyan, Qian Binghao

Figure 1 The typical example of image processing Lena ——Lena Soderberg，born in Mar.31st in 1951 in Sweden. She acted as cover girl of the Nov. edition of playboy, 1951. The mid page photograph was photo’d by Dwight Hooker, which was widely used in the Image Processing in the later decades.

Abstract: Through the inspiration of Dean Zen’s lecture, this paper aims to clearly interpret the process of Hill’s cipher and the mathematical method in it, also we will sequentially focus on the deeper use of Hill’s cipher as well as applications, and eventually deduce the procedure with the help of MATLAB.

Key Words: Hill cipher, Cryptography, inverse of the matrix, decryption, encryption, image encryption.

1. Introduction

It’s generally accepted by all that network and electronic commerce have played vital role in nowadays individuals daily life due to the rapid development of computer technology and modern times communication technique. Digitization as well as informatization are tend to be the trend and flourish of future technology development.

However, with the widely use of Internet, some crimes and frauds are conducted by using the leak of information and unsafe security cyber surroundings, so it is of critical significance to pay attention to the information security. According some reference book, we’ve already know that the modern field of cryptography are divided into several areas of study such as symmetric-key cryptography, public-key cryptography as well as some encryption key.

This paper is mainly focus on one type of cryptography named Hill cipher, which is a classical polygraphic substitution cipher based on linear algebra. This kind of cipher was invented by Lester. S. Hill in 1929.We’ll profoundly describe the theory of the Hill cipher by using some basic knowledge of elementary matrices (inverse matrices), and with the help of MATLAB and using some examples to introduce the further use of Hill cipher as well as other different kinds of way of cryptograph.

2. Section 2

2.1

The basic method of Hill cipher is using a key matrix to multiply a plaintext vector in order to get a ciphertext vector and change it into the string of ciphertext. The key matrix is actually a transformation matrix. We suppose the plaintext vector is

P=p1p2⋮pn with 26 characters changed into numbers by using the ASCII（American Standard Code for Information Interchange）,while the key matrix is a n*n matric matrix K= k11⋯k1n⋮⋱⋮kn1⋯knn .Also suppose the ciphertext vector is C=c1c2⋮cn , then use the equation:

C=KP (mod26)

to get the ciphertext vector:

c1=(k11p1+k12p2+k13p3+…+k1npn)mod26c2=k21p1+k22p2+k23p3+…+k2npnmod26⋮cn=(kn1p1+kn2p2+kn3p3+…+knnpn)mod26

And we will decrypt the cipher by using the inverse of the matrix like:

P=K-1C(mod26)

In which KK-1=I (I has to be an identity matrix)

2.2

In order to profoundly explain the procedure of decryption as well as encryption, we take a convincing example to illustrate the process of Hill’s cipher.

Encryption

1)We transform Chinese character into Chinese phonetic alphabet as a Plaintext:

Example:

you nei gui

Divide the plaintext into n-letter groups (n is determined by the key)

you nei gui

2)Each plaintext letter is assigned a numerical value from 0 to 25, similarly as in the Caesar cipher from a to z.

a	b	c	d	e	f	g	h	i	j	k	l	m
0	1	2	3	4	5	6	7	8	9	10	11	12
n	o	p	q	r	s	t	u	v	w	x	y	z
13	14	15	16	17	18	19	20	21	22	23	24	25

Plaintext: y o u n e i g u i

24 14 20 , 13 4 8, 6 20 8

3)Fill the plaintext values into the P matrix

(from peak to bottom, from left to right)

P=24136144202088

4)Use the encryption algorithm:
C=KP (mod26)
(Point: The matrix multiplication is not commutative KP and PK are not equal in general)

C=k11⋯k1n⋮⋱⋮kn1⋯knnp11⋯p1n⋮⋱⋮pn1⋯pnn(mod26)

cij=k=1nkikpkj

C=c11c21⋯cn1⋯cnn

In this case, we set the key
K=171752118212219 and P=24136144202088

c11=(k11p11+k12p21+k13p31)mod26c21=(k21p11+k22p21+k23p31)mod26c31=(k31p11+k32p21+k33p31)mod26

C=181714619414422

Which use the Caesar list to transform it and result in sgo rte oew

Decryption

The process of decryption involves the inverse of the matrix

K-1K=I ,IP=P

C=EK,P=KP mod 26

P=DC,P=K-1C mod 26

=K-1KP=IP=P

So, the key factor to get P is to calculate K-1C mod 26

Before we start to do calculation, we shall check the Modular Arithmetic

a+bmod m=[a mod m+b mod m]mod m

a-bmod m=[a mod m-b mod m]mod m

a×bmod m=[a mod m×b mod m]mod m

Which means
P=K-1Cmod26=[ K-1 mod 26×C mod 26]mod 26

K-1mod 26={[det K-1 mod 26]×K* mod 26]}mod 26

=-939-1×300-303267-357313-25260-51

=23-1×14257718601

At this step we should apply Modular Multiplicative Inverse

There exist a∈Zm,a-1∈Zm,

Which a a-1=1 mod m

(** Not every integer a has a multiplication inverse, it exists if and only if a and m are coprime)

The modular multiplicative inverse of x modulo y can be found with the extended Euclid algorithm if

ax+by=1

Then

y-1=b mod x

a	1	3	5	7	9	11	15	17	19	21	23	25
a-1 mod 26	1	9	21	15	3	19	7	23	11	5	17	25

K-1mod 26={[det K-1 mod 26]×K* mod 26]}mod 26

=-939-1×300-303267-357313-25260-51mod 26

=23-1×14257718601mod 26

=17×14257718601mod 26

=23842511911917136102017mod 26

=49151517624017

Then we apply this into P=K-1Cmod26

=49151517624017181714619414422mod 26

=336299422456602410670476710mod 26

=24136144202088

Then we transform the value into characters and result in the ciphertext:

you nei gui

3. Section 3

This part introduces a general process applying Hill cipher in image encryption, whose main purpose is covering the information contained in the image,

Matrix Model

2×2 matrix model

Every pixel in a image is formed by three kinds of RGB color value(R-Red, G-Green, B-Blue). And each value will be expressed as a integer from 0 to 255 in computer. Because of that, we can put RGB values of each pixel with another value to constitute a 2×2 matrix whose elements are the values above. It can be expressed like Pij=Xij1Xij2Xij3Xij4 and the sequence of Xij1,Xij2,Xij3,Xij4 can be an arbitrary sequence of Rij,Gij,Bij,Eij . (The arrangement can also be a part of the “plaintext”)

*The options of Eij are various. But it must satisfy the condition of being a integer from 0 to 255, which can be the digital signature, the random numbers or even a constant.

3×1 matrix model

We can just simply put three RGB values into a 3×1 matrix like Pij=Xij1Xij2Xij3 (the sequence of Xij1,Xij2,Xij3 can be an arbitrary sequence of Rij,Gij,Bij and the arrangement can also be a part of the “plaintext”).

Logistic Map

Logistic map is a kind of dynamic system which is widely studied. The definition of it is

Xn+1= λXn(1-Xn), n=0,1,2…

As a phenomenon found in non-linear dynamic system, chaos is a deterministic and random-like process. Researches shows that when parameter λ∈(3.5699456,4), logistic map is at chaos state. Therefore, the resulting chaotic sequence is suited for generating encryption sequence.

Encryption

Ⅰ. Determine the matrix size and arrangement of elements

2×2 matrix model

Let

Kij=rijrij+1rij-1rij, rij∈0,255∩Z

, then

K=1

(1 is a coprime factor of 256).

For convenience, let

Pij=RijGijBijEij

Cij=RijGijBijEij=KijPij(mod256)= Rijrij+Bijrij+BijGijrij+Eijrij+EijRijrij+Bijrij-RijGijrij+Eijrij-Gij(mod256)

, and the Rij , Gij , Bij are the new RGB values of every pixel.

3×1 matrix model

Let

Kij=sijtijtijtijsijtijtijtijsij, sij, tij∈0,255∩Z, sij≠tij

(otherwise the matrix will not be invertible)

, then

K=(sij+2tij)(sij-tij)2

((sij+2tij ) and (sij-tij )must be coprime factors of 256).

For convenience, let

Pij=RijGijBij, Cij=RijGijBij=KijPij(mod256)=Rijsij+Gijtij+BijtijRijtij+Gijsij+BijtijRijtij+Gijtij+Bijsij(mod256)

.(and the Rij , Gij , Bij are the new RGB values of every pixel).

II. Generate encrypting matrices

Setup parameter λ∈(3.5699456,4) and initial value X0 ∈[0,1] for the guarantee of Xn ∈[0,1], then a chaotic sequence is got. Let

rij=[255Xi-1+j]

([x] is the rounding function, x-1<[x]≤x),

sij=2[127Xi-1+j]+1&tij=2[128Xi-1+j]

(to satisfy that (sij+2tij ) and (sij-tij ) are coprime factors of 256)

III. Compute

Using formula compute every pixel’s color after encryption.

Decryption

I. Confirm the secret key

Confirm all the information for decryption included the arrangement of pixel matrix, λ and X0 .

II. Generate Decryption matrices and Compute

For 2×2 encryption matrices, use formula

Pij=RijGijBijEij=Kij-1Cijmod256=RijGijBijEijrij-rij-1-rij+1rijmod256

(The Kij-1 =1 and the modular multiplicative inverse of 1 modulo 256 is also 1.)

to restore the original image.

While for 3×3 encryption matrices, use following formulas.

(Rij+Gij,+ Bij)mod256

=(Rijsij+Gijtij+Bijtij)+( Rijtij+Gijsij+Bijtij)+( Rijtij+Gijtij+Bijsij)mod256

=(sij+2tij)( Rij+Gij+Bij)mod256=(sij+2tij)Kijmod256

(Kij=Rij+Gij+Bij , and Kij is the only solution due to sij+2tij is a coprime factor of 256)

Rij=Rijsij+Gijtij+BijtijKijtij=Rijtij+Gijtij+Bijtij⇒(Rij-Kijtij)mod256=Rij(sij-tij)mod256

(Rij is the only solution due to sij-tij is a coprime factor of 256, as well as Gij and Bij )

(a).Original image

(b). 2×2 matrix encrypted image (c). 3×1 matrix encrypted image

(d). 2×2 matrix decrypted image (e). 3×1 matrix decrypted image

4. Section 4

This part provides two sample MATLAB codes for encryption and decryption respectively.

NB the program runs on Capitalized information and only transfer letter in the unit of a certain length of combination which depends on the dimension of key matrix.

Part I Encryption

function hill()
%Enter key
disp('Enter the dimension of the key matrix');
n=input('');
disp('Enter the key matrix in the order of rows');
key=zeros(n,n);
for j=1:n;
    for k=1:n
        key(j,k)=input('');
    end
end
d=det(key);%Determinant
if d==0 
   error('Not invertible.Cant hill');
end 
%Enter info
message=input('Enter the inf in CAPITALS \n','s');
m=size(message);
m=m(2);%Get length
if mod(m,n)~=0
    error('Error.Entering code should be multiplier of the key matrix dimension');
end
for i=1:m
    if message(i)>='A' && message (i)<='Z'  
        message(i)=message(i)-65;
    else 
        error('Error. Captial letters are needed.');
    end 
end
%Encryption
i=1;
while i<m+1
    B=message(i:i+n-1)';
    a=key*B;
    A(i:i+n-1)=a';
    i=i+n;
end
for i=1:m
    if A(i)>26
        A(i)=mod(A(i),26);
    end 
    A(i)=A(i)+65;
end
str=char(A);
fprintf('The code is %s',str)
end

Part II Decryption

%Decryption
function hill2()
%Enter key
disp('Enter the dimension of the key matrix');
n=input('');
disp('Enter the key matrix in the order of rows');
key=zeros(n,n);
for j=1:n
    for k=1:n
        key(j,k)=input('');
    end
end
d=det(key);%Determinant
 
if d==0
    error('Not invertible.Cant hill');
end
%ÊäÈëÃÜÎÄ
message=input('Enter the code\n','s');
m=size(message);
m=m(2);
if mod(m,n) ~=0
    error ('Error.Entering code should be multiplier of the key matrix dimension');
end 
for i=1:m
    if message(i)>='A'&& message(i)<='Z'
        message(i)=message(i)-65;
    else 
        error('Error. Captial letters are needed.');
    end 
end 
%r1 is the modular multiplicative inverse of d 
while abs(d)>26
    d=mod(d,26);
end
switch d 
    case 1
        r1=1;
    case 3 
        r1=9;
    case 5
        r1=21;
    case 7
        r1=15;
    case 9
        r1=3;
    case 11
        r1=19;
    case 15
        r1=7;
    case 17
        r1=23;
    case 19
        r1=11;
    case 21
        r1=5
    case 23
        r1=17
    case 25
        r1=25;
    otherwise disp('The reciprocal of d doesnt exist');
end 
detk=det(key);
invk=inv(key);
k=detk*invk;
key2=mod(r1*k,26);
for i=1:n
    for j=1:n
        if key2(i,j)<0
        key2(i,j)=key2(i,j)+26;
        end 
    end
end 
i=1;
while i<m+1
    p= message(i:i+n-1)';
    a=key2*p;
    B(i:i+n-1)=a';
    i=i+n;
end
for i=1:m
    if B(i)>26
        B(i)=mod(B(i),26);
    end
    if B(i)==0
        B(i)=26;
    end
    B(i)=B(i)+65;
end
str2=char(B);
fprintf('\nThe inf is %s\n',str2);
end

The command window result screenshots:

Figure 1 Figure 2

5. Conclusions

This paper meanly illustrate the acute procedures of encryption and decryption of Hill’s cipher by applying the knowledge of inverse of matrix and modular arithmetic. Additionally, it shows the deeper use of Hill’s cipher on image, text and other applications which plays an essential part in nowadays network security and electronic commerce even in the military fields.

References

Xu Xiao-hua and Li Min-ying(2010), The Solution of Matrix of Hill Cipher in Encryption and Decryption.
Ahmed Y. Mahmoud, Alexander G. Chefranov(2014), A Hill Cipher Modification Based on Eigenvalues Extension with Dynamic Key Size HCM-EXDKS.
Liu jiwei(2010), Digital image encryption algorithm based on hill cipher.