JOURNAL OF COMPUTING, VOLUME 4, ISSUE 8, AUGUST 2012, ISSN (Online) 2151-9617 https://sites.google.com/site/journalofcomputing WWW.JOURNALOFCOMPUTING.

ORG

115

Design and Implementation of Network Security using Neural Network Architecture

1
1

Inadyuti Dutt, 2Soumya Paul, 3Sudipto Chandra

3

Assistant Prof., Assoc. Prof., Dept. of Computer Application, B. P. Poddar Institute of Management & Technology, Affiliated to West Bengal University of Technology

MCA final year Dept. of Computer Application, B. P. Poddar Institute of Management & Technology, Affiliated to West Bengal University of Technology

Abstract: Over the last few years, secured transmission of data has been a major issue in data communication. This implementation mainly concerns about the security of confidential information and data transmission through Neural Network in order to provide confidentiality, authentication, integrity and non-repudiation of the messages. First, an encryption algorithm is developed and implemented to achieve the aforesaid purpose. It is basically a program that takes any plain text as input from the user and produces a cipher text which is sent to the source node of neural network. Now the cipher text is again encrypted and fragmented in successive layers of neural networks concept and finally all the fragments are collected and decrypted at destination node to get the cipher text which was input to the source node. Then the cipher text is decrypted to get original plain text.

Keywords: Neural Network (NN)

1. Introduction:
Network security consists of the provisions and policies to prevent and monitor unauthorized access, misuse, modification, or denial of a computer network and network-accessible resources. Network security involves the authorization of access to data in a network, which is controlled by the network administrator. Users choose or are assigned a password or other authenticating information that allows them access to information and programs within their authority. Network security covers a variety of computer networks, both public and private, that are used in everyday jobs conducting transactions and communications among businesses, government agencies and individuals. based on external or internal information that flows through the network during the learning phase. Modern neural networks are non-linear statistical data modeling tools. They are usually used to model complex relationships between inputs and outputs or to find patterns in data.

1.2 Outline of Remaining Section:

The paper is organized as follows: Section 2 states the problem and the security of files that contain certain confidential information. The detailed description of the proposed encryption decryption algorithms with illustration of examples are explained in section 3. Finally, the paper concludes in section 4.

2. Problem Statement
Encryption method follows as the first and the basic criteria in order to encrypt a plain text is to input a password because the key array is generated from this password. In symmetrickey cryptography, it is obvious that when a text is encrypted with a secret key, it is decrypted with the same key. The plain text is encrypted to produce cipher (cipher 0) using key array. The cipher 0 text is sent as input to the source node of neural network and the cipher get encrypted in successive layers and

1.1. Neural network:

A neural network consists of an interconnected group of artificial neurons, and it processes information using a connectionist approach to computation. In most cases an ANN is an adaptive system that changes its structure

2012 Journal of Computing Press, NY, USA, ISSN 2151-9617

JOURNAL OF COMPUTING, VOLUME 4, ISSUE 8, AUGUST 2012, ISSN (Online) 2151-9617 https://sites.google.com/site/journalofcomputing WWW.JOURNALOFCOMPUTING.ORG

116

decrypted at destination node to produce the initial cipher (cipher 0). The same key array is used in decryption to get back the actual text. So, decryption requires knowledge of the same key.

Send e1 to NODE N+1 of layer L+1 ELSE Step 4: Set e: = ASCII (cipher character) key[i]-w. Send e to Node N +2 of layer L+1 Step 5: Set i: = i+1. Step 6: Stop

Encryption: cipher 0 = encrypt (plaintext, key []) Encryption at layer l of neural network: cipher n= decrypt (cipher, key [], weight of node n ) Decryption at destination: cipher 0 = decrypt ( cipher n, key[] ,weight of destination) Decryption: Plaintext = decrypt (cipher 0, key [])

3.3 Proposed Decryption Heuristic Using Neural Network

Input: Encrypted Cipher text from layer L-1 Output: Decrypted Cipher Step 1: Start Step 2: Set weight w: =x, i=1 Step 3: if cipher character pos is Odd then Set d: = ASCII (encrypted cipher) -key[i]-w ELSE Step 4: Set d: = ASCII (encrypted cipher) +key[i] +w Step 5: Set i: = i+1 Step 6: Stop

3.4 Proposed Decryption Heuristic

Step 1: Start Step 2: Take all alphabet from A to Z in a matrix I (6X6). Step 3: Take a password Step 4: Reverse the row and column of the matrix I (6X6) according to the password to get matrix R (6X6). Step 5: Take the Decrypted text in a square matrix A Step 6: Reverse the Even column and Odd row of matrix A respectively Step 7: Decrypt each element of matrix A using position of elements of matrix I and R to generate plain text. Step 8: Stop

3. Proposed Algorithm 3.1. Proposed Encryption Heuristic

Input: Plain text is given. Output: Encrypted cipher text. Step 1: Start Step 2: Take all alphabet from A to Z in a matrix I (6X6). Step 3: Take a password Step 4: Reverse the row and column of the matrix I (6X6) according to the password to get the matrix R (6X6). Step 5: Take the plain text in a square matrix A Step 6: Reverse the Odd row and Even column of matrix A respectively. Step 7: Encrypt each element of matrix A using position of elements of matrix I and R. Step 8: Stop

3.5 Proposed Cryptographic Algorithm

Step 1: Start Step 2: Call Proposed Encryption Heuristic//Subsection 3.1 Step 3: Call Proposed Encryption Heuristic using Neural Network//Subsection 3.2 Step 4: Call Proposed Encryption Heuristic using Neural Network//Subsection 3.3 Step 5: Call Proposed Decryption Heuristic Step 6: Stop

3.2 Proposed Encryption Heuristic Using Neural Network

Input: Cipher text obtained in initial encryption, List of keys key [] from password Output: Successively encrypted Cipher text. Step 1: Start Step 2: Set weight w: = x, i: =1 Step 3: IF cipher character pos is Odd then Set e: = ASCII (cipher character) +key[i] +w

3.6 Illustration of Proposed Algorithm with example A. Encryption

2012 Journal of Computing Press, NY, USA, ISSN 2151-9617

JOURNAL OF COMPUTING, VOLUME 4, ISSUE 8, AUGUST 2012, ISSN (Online) 2151-9617 https://sites.google.com/site/journalofcomputing WWW.JOURNALOFCOMPUTING.ORG

117

Sep 1: We take @ = 0 and [A-Z] = [1-26] as follows: 0 @ 1 A 2 B 3 C 4 D 5 E 6 F 7 G 8 H 9 I 10 J 11 K 12 L 13 M 14 N 15 O 16 P 17 Q 18 R 19 S 20 T 21 U 22 V 23 W 24 X 25 Y 26 Z

Now we take an Initial Matrix I (6x6):

0 1 2 3 4 5 0 @ K L R X 1 A J M S Y 2 B I N T Z 3 U O H C 4 D G P V 5 E F Q W

STEP 2: We take a password: HONEYBEE

H 8 18 O 15 11 N 14 12 E 5 21 Y 25 1 B 2 24 E 5 21 E 5 21

(pos) (26pos)

Continuing in similar way, we finally get the following:

0 1 2 3 4 5 0 Z E Q K @ 1 A J P V Y 2 X T H C B 3 U O N I 4 D G M S 5 W L F R

So, We take Key[ ]={18,11,12,21,1,24,21,21} STEP 3: We change our Initial Matrix I (6 X6) into a new Reference Matrix R (6X6) using position of element of password (pos):

This is our Reference Matrix R (6X6) STEP 4: Let us take a plain text which is required to be encrypted. Plain text: THE SKY IS BLUE The text consists of 15 characters. So, a (4X4) matrix would be sufficient to accommodate all characters as follows:
0 T S I L 1 H K S U 2 E Y @ E 3 @ @ B @

The 1st elemen t in password array is H. So, we reverse the Row that consists of H and we get the following:
0 @ F L R X 1 A G M S Y 2 B H N T Z 3 C I O U 4 D J P V 5 E K Q W

0 1 2 3 4 5

0 0 6 12 18 24

1 1 7 13 19 25

2 2 8 14 20 26

3 3 9 15 21

4 4 10 16 22

5 5 11 17 23

0 1 2 3 4 5

0 1 2 3

0 1 2 3
@ K L R X A J M S Y B I N T Z C H O U D G P V E F Q W

0 20 19 9 12

1 8 11 19 21

2 5 25 0 5

3 0 0 2 0

0 1 2 3 4 5

STEP 5: Now, we take a matrix B (4X4) where B (i, j) =26-A (i, j). Matrix (B) =>
0 1 2 3 0 6 7 17 14 1 18 15 7 5 2 21 1 26 21 3 26 26 24 26

The 2nd element in password array is O. So, we reverse the Column that consists of O and we get the following:

2012 Journal of Computing Press, NY, USA, ISSN 2151-9617

JOURNAL OF COMPUTING, VOLUME 4, ISSUE 8, AUGUST 2012, ISSN (Online) 2151-9617 https://sites.google.com/site/journalofcomputing WWW.JOURNALOFCOMPUTING.ORG

118

G
0 F G Q N 1 R O G E 2 U A Z U 3 Z Z X Z

0 1 2 3

STEP 6: Now we reverse each ODD Row of Matrix B

0 26 7 24 14 1 21 15 26 5 2 18 1 7 21 3 6 26 17 26

0 1 2 3

0 1 2 3

0 Z G X N

1 U O Z E

2 R A G U

3 F Z Q Z

STEP 7: Now we reverse each EVEN Column of Matrix B

0 26 7 24 14 1 5 26 15 21 2 18 1 7 21 3 26 17 26 6

=> 7 => I (1, 1) => R (1, 1) => J Z => 26 => I (4, 2) => R (4, 2) => B A => 1 => I (0, 1) => R (0, 1) => A Q => 17 => I (2, 5) => R (2, 5) => F X => 24 => I (4, 0) => R (4, 0) => @ O => 15 => I (2, 3) => R (2, 3) => N G => 7 => I (1, 1) => R (1, 1) => J Z => 26 => I (4, 2) => R (4, 2) => B N => 14 => I (2, 2) => R (2, 2) => H U => 21 => I (3, 3) => R (3, 3) => I U => 21 => I (3, 3) => R (3, 3) => I F => 6 => I (1, 0) => R (1, 0) => E So, the text now becomes: BWKBJBAF@NJBHIIE i.e. BWKBJBAF NJBHIIE which is to be send through neural network.

0 1 2 3

B. Neural Network
Text received at Node 1: BWKBJBAF@NJBHIIE Node 1: Weight (w) =1 Read the 1st character: B position (1st position)] ASCII (B) =67 Weight (w) =1 1) Key [0] =18 ASCII (66-18-

[B is in Odd

0 1 2 3

0 Z G X N

1 E Z O U

2 R A G U

3 Z Q Z F

The sentence becomes: ZERZGZAQXOGZNUUF STEP 8: Now, we encrypt the above text taking the position of each character in the text with respect to our Initial Matrix I (6 X6) and corresponding Reference matrix R (6X6) as shown below: Z => 26 => I (4, 2) => R (4, 2) => B E => 5 => I (0.5) => R (0, 5) => W R => 18 => I (3, 0) => R (3, 0) => K Z => 26 => I (4, 2) => R (4, 2) => B

ASCII (47) / Node 3

Node 1: Weight (w) =1 Read the 2nd character: W [W is in Even position (2nd position)] ASCII (w) =87 Weight (w) =1 Key [1] =11 ASCII (87+11+1) ASCII (99) Node 2 c

2012 Journal of Computing Press, NY, USA, ISSN 2151-9617

JOURNAL OF COMPUTING, VOLUME 4, ISSUE 8, AUGUST 2012, ISSN (Online) 2151-9617 https://sites.google.com/site/journalofcomputing WWW.JOURNALOFCOMPUTING.ORG

119

And so on.

C. Decryption
STEP 1: We take @ =0 and [A-Z] = [1-26] as follows: 0 @ 10 J 20 T 1 A 11 K 21 U 2 B 12 L 22 V 3 C 13 M 23 W 4 D 14 N 24 X 5 E 15 O 25 Y 6 F 16 P 26 Z 7 G 17 Q 8 H 18 R 9 I 19 S

0 1 2 3 4 5

0 Z E Q K @

1 A J P V Y

2 X T H C B

3 U O N I

4 D G M S

5 W L F R

STEP 4: Text received at Node 6(Destination Node): BWKBJBAF@NJBHIIE Now, we decrypt the above text taking the position of each character in the text with respect to our Initial Matrix I (6 X6) and corresponding Reference matrix R (6X6) as shown below: B => => => => => => => => => => => => => => => => => => => => => => => => => => => => => => => => R (4, 2) => 26 => R (0, 5) 5 R (3, 0) 18 R (4, 2) 26 R (1, 1) 7 R (4, 2) 26 R (0, 1) 1 R (2, 5) 17 R (4, 0) 24 R (2, 3) 15 R (1, 1) 7 R (4, 2) 26 R (2, 2) 14 R (3, 3) 21 R (3, 3) 21 R (1, 0) 6 => => => => => => => => => => => => => => => => => => => => => => => => => => => => => => I (4, 2) Z I (0.5) E I (3, 0) R I (4, 2) Z I (1, 1) G I (4, 2) Z I (0, 1) A I (2, 5) Q I (4, 0) X I (2, 3) O I (1, 1) G I (4, 2) Z I (2, 2) N I (3, 3) U I (3, 3) U I (1, 0) F

Now we take an Initial Matrix I (6x6):

W
0 1 2 3 4 5 0 0 6 12 18 24 1 1 7 13 19 25 2 2 8 14 20 26 3 3 9 15 21 4 4 10 16 22 5 5 11 17 23

Now w

K B J B

STEP 2: We take a password: HONEYBEE H 8 O 15 N 14 E 5 Y 25 B 2 E 5 E 5

A F @

(pos)

STEP 3: We change our Initial Matrix I (6 X6) to generate a new Reference Matrix R (6X6) using position of element of password (pos) as shown in Step 3 of Encryption.

N J B H

0 1 2 3 4 5

0 @ F L R X

1 A G M S Y

2 B H N T Z

3 C I O U

4 D J P V

5 E K Q W

I I E

Taking the characters in Matrix a (4 X 4)

Reference Matrix R (6X6):

0 1 2 3 0 Z G X N 1 E Z O U 2 R A G U 3 Z Q Z F

2012 Journal of Computing Press, NY, USA, ISSN 2151-9617

JOURNAL OF COMPUTING, VOLUME 4, ISSUE 8, AUGUST 2012, ISSN (Online) 2151-9617 https://sites.google.com/site/journalofcomputing WWW.JOURNALOFCOMPUTING.ORG

120

0 1 2 3

0 26 7 24 14

1 5 26 15 21

2 18 1 7 21

3 26 17 26 6

STEP 5: Now we reverse each EVEN Column of Matrix A

0 Z G X N 1 U O Z E 2 R A G U 3 F Z Q Z

public and ii) Private Key, which is kept secret. Any message encrypted using private key can be decrypted using public key or vice versa. So, the message gets decrypted when it is encrypted by only the authenticated person who knows the private key.

5. References: Books:
1. Cryptography and Network Security, Third Edition, by William Stallings. Cryptography And Network Security (Sie), By Forouzan. An Introduction to Neural Networks, by Kevin Gurney, Kevin N. Gurney.

0 1 2 3

0 1 2 3

0 26 7 24 14

1 21 15 26 5

2 18 1 7 21

3 6 26 17 26

2. 3.

STEP 6: Now we reverse each ODD Row of Matrix A

0 F G Q N 1 R O G E 2 U A Z U 3 Z Z X Z

0 1 2 3

0 1 2 3

0 6 7 17 14

1 18 15 7 5

2 21 1 26 21

3 26 26 24 26

STEP 7: Now, we take a matrix B (4X4) where B (i, j) =26-A (i, j)

0 1 2 3 0 20 19 9 12 0 T S I L 1 8 11 19 21 1 H K S U 2 5 25 0 5 2 E Y @ E 3 0 0 2 0 3 @ @ B @

0 1 2 3

So, finally the decrypted text: THE SKY IS BLUE.

4. Conclusion:
Implementation of Cryptographic Algorithm (encryption and decryption Schemes) using asymmetric keys contains two keys: i) Public Key, which is known to

2012 Journal of Computing Press, NY, USA, ISSN 2151-9617