A Novel DNA Encryption System Using Cellular Automata

International Journal of Security, Privacy and Trust Management (IJSPTM) Vol 4, No 3/4, November 2015
DOI : 10.5121/ijsptm.2015.4404 39
A NOVEL DNA ENCRYPTION SYSTEM USING
CELLULAR AUTOMATA
G.Shanmugasundaram1
, P.Thiyagarajan2
, S.Pavithra1
1
Department of Information Technology, Sri Manakula Vinayagar Engineering College,
Madagadipet, Puducherry.
2
Department of Computer Science and Engineering, Mahindra Ecole Centrale, College of
Engineering, Hyderabad - 43.
ABSTRACT
DNA Cryptography is a new born cryptographic field emerged with the research of DNA Computing in
which DNA is used as an Information carrier. Cellular automata is dynamic in nature so it provide
dynamic behavior in the system which may increase the security in the system. DNA cryptography is
provide a secure way to encrypt the text and automata changes the state of the system based on the present
state, it will occur in discrete time. These qualities are most impressive in these techonology which help us
to provide a highly secured security system for the users. Most of the encryption techniques based on the
cellular automata have limitations. To overcome this lacuna, we propose a novel DNA cryptography
algorithm with cellular automata to achieve randomness, parallelism, uniformity, reversibility and stable.
An algorithm implemented and its results obtained are depicts here, and a result analysis is done with
other algorithms.
KEYWORDS
Cellular automata, DNA, Thymine, Uracil.
1. INTRODUCTION
Nowadays, Internet has become parts and parts of human life and it has emerged closely with
human such that it is elevated as one of the basic needs of human life. The facility of using the
Internet in mobile phones has increased its use among common man. Many sectors of business
such as bank, travel, shopping etc., make use of Internet to reach many people and flourish their
business by introducing their application (apps). Though there are many advantages of Internet, it
has adverse effect such as security. Sensitive information such as financial transactions,
confidential government information, and medical records is transmitted through Internet. The
protection of this sensitive information in the Internet is done by information security systems
such as cryptography [1], steganography [2] [3] and combination of cryptography and
steganography [4]. Security in Internet is provided by the information security systems. With the
advancement in computational infrastructure and computer processing capacity, standard
information security systems are no longer secure1
. As traditional cryptographic methods are
based on mathematical and theoretical models, with the recent advancements in computational
techniques, cryptographic methods are vulnerable to attacks. A novel technique for securing data
was introduced using the biological structure of DNA called DNA Computing. It was invented by
Leonard Max Adleman in the year 1994, for solving the complex problems such as directed
Hamilton path problem, NP-complete problem and travelling salesman problem.
1
http://www.symantec.com/connect/blogs/2015-internet-security-threat-report-attackers-are-bigger-bolder-and-faster

40
Fig.1. Cell Definition
The DNA computing is also used in cryptography research. DNA is used in cryptography for
storing and transmitting the information, as well as for computation. This paper, try to utilize the
power of DNA to hide message [5]. The massive parallelism and extraordinary information
density inherent in this molecule are exploited for cryptographic purposes. The current main
difficulties of DNA cryptography are the absence of theoretical basis, the high tech lab
requirements and computation limitations [6]. In this paper DNA cryptography is used along with
cellular automata. A brief introduction of cellular automation is described in the below paragraph.
Cellular automation is a system made up of many discrete cells, each of which may be in one of a
finite number of states [16]. A cell or automaton may change state only at fixed, regular intervals,
and only in accordance with fixed rules [20] that depend on cells own values and the values of
neighbors within certain proximity. The formal definition of Cellular Automation (CA) is one or
two dimensional grid [10] of identical automata cells. Each cell processes information and
proceeds in its action depending on its neighbors. Each cell (automaton) is defined by
• Set of States S = {S1, S2, S3,S4…, SN}
• Transition Rules T
• Therefore A ~ (S,T,R) (R: neighboring automata)
• T: (St, It) → St+1
The rest of the paper is organized as follows: Section 2 of this paper deals with related work in
DNA cryptography. The proposed work and experiment methodology are discussed in section 3.
Experiment results of the proposed work are presented in section 4. Section 5 concludes the
paper.
2. LITERATURE SURVEY
This section describes some of the DNA computing and cellular automata methods reported in
literature and it also describes the motivation to combine DNA computing and cellular automata.
According to Guangzhao Cui et al [8], DNA computing is a new computational paradigm by
harnessing the potential massive parallelism, high density information of bio-molecules and low
power consumption, which brings potential challenges and opportunities to traditional
cryptography. Guangzhao Cui et al [8] in his paper discussed about DNA computing and its
application in information security field and reviews the principle of DNA computing. It also
elaborates about schemes with secret key searching and introduces the application of DNA
computing in encryption, steganography and authentication [8].
S.Jeevidha et al [21] proposed an analysis on DNA based cryptography to transfer the data
securely and also discusses DNA technology used in the cryptographic methods. DNA
technology includes all biological process like polymerase chain reaction and electrophoresis.

41
H.Z.Hsu [7] in his method exploits the interesting properties of DNA sequence to encrypt data.
The author presents three methods namely, the insertion method, the complementary pair method
and the substitution method. In [8] the paper analyse some schemes with secret key searching and
introduce the application of DNA computing in encryption, steganography and authentication.
Anupriya et.al [9] in her paper proposed an algorithm which is based on reference sequence
known only to sender and receiver. Since there are roughly 55 million publicly available DNA
sequence it is virtually impossible to crack the sequence.
A Study of Data Encryption Standard (DES) Algorithm with Cellular Automata was proposed by
Rama R et al [11]. DES algorithm was accepted as the standard encryption algorithm in 1976.
Since then the development and usage of DES is widely recognized. The concept of Cellular
Automata (CA) introduced by John von Neumann (1950s) is the first model which exhibited the
capability of producing complex and random behavior.
Cellular Automata is a parallel processing machine and it involves context dependent behavior of
all the value at time ‘t’ [17]. The literature has witnessed the usage of Cellular Automata rules
which are simple in nature for the operation of DES and Advanced Encryption Standard (AES)
algorithms.
Cellular automata is applied in variety of fields such as smart devices to provide security [14],
pattern classification [15], traffic modelling etc. Rama R et al [11] extend the usage of Cellular
Automata (CA) for the key generation in DES. The proposed CA-based key generation
methodology uses the elementary Cellular Automata rule 30 which possess more randomness.
The initial seed involved is 15 bit multi-seed and presented the implemented DES with key
generated using Cellular automata in Java. Due to cellular automata rule 30 randomness is highly
provided in this method. The demerit of this method is the key size is minimal in this approach.
Pratibha Sharma et al [12] proposed Text Security Using 2D Cellular Automata Rules Encryption
method which is the most common method for hiding text from unauthorized access. Encryption
provides only one level of security during transmission over the channel. The main aim would be
to provide 2 levels of security. First level comprises of hiding text to be sent behind some image
using password and the second level comprises of encryption using 2D Cellular rules. If one level
security is broken then the other level would provide security thereby ensuring more security to
the transmitted message. Encryption would be done using 2 dimensional rules of Cellular
Automata. Moore neighborhood model have been used as Cellular Automata model. In Moore
neighborhood model 9 cells are considered at a time including the cell itself. The value of current
cell (i.e. central cell) depends on its 8 neighbors. The merits of this method are: two level
securities are provided in this method and the other is due to chip in automata machine which
enhances the speed of encryption and decryption.
Triple Stage DNA Cryptography Using Sequential Machine was proposed by Rupali Soni et al
[13], to secure data there are some traditional cryptographic techniques such as substitution,
transposition, and various algorithms such as RSA, DES, AES, IDEA etc. and protocols like SSL
are used. This work in [13] represents a new approach to secure the data which is based on DNA
logic, cryptographic scheme and automata theory. DNA Cryptography technique was proposed
by L. M. Adleman in 1994 for solving Hamiltonian path problem and automata theory was firstly
initiated in 20th century. In the proposed work [13] these two techniques are combined for
creating powerful security algorithm. This algorithm is more reliable and powerful because of the
use of secret key, auto generated Moore machine and use of password [13]. The merits of the
method are security is provided in triple stage in this method and it is resist to many known
attacks.

2.1. Merging of DNA Cryptography and Cellular Automata
DNA cryptography and cellular automata used in the cryptography field, but both are not
combined to provide security. As both domains are yielding better solutions to the security issues,
if these fields are combined, security provided by them will be enha
this paper combines both fields in an efficient way for developing highly secured system. As
cellular automata are dynamic in nature it contributes for dynamic behavior in the system which
increases the overall security. DNA cryp
These qualities are most impressive which motivates to provide a highly secured security system
for the users. The ultimate aim of this proposed work is to provide high level security which may
make the hackers less possible to crack the system.
3. PROPOSED SYSTEM
The objective of the work is to develop a prominent secure system that satisfies the needs of the
user. The DNA cryptography and cellular automata will lead to provide effective cryptographic
algorithms. Each field has their own characteristics to establish if they are combined to form a
cryptosystem it will lead to achieve properties of both fields.
technique of both DNA cryptography and cellular automata and it is
advanced encryption algorithm using DNA substitutaion combined with Elementary Cellular
automata rule (ECA-51)” [18] [19].
3.1. Working of proposed system
Initially the plain text is taken from the user either in file or direct input.
convert the data in the file to DNA sequence until the end of file. Then convert the DNA
sequence into binary file and divide the binary sequences into parts and perform xor operation on
the binary parts with the help of key on
of xor operation i.e binary sequence into DNA sequence and send to automata machine. Here
with the help of automata machine the DNA sequence is convert into DNA code words which is
in the form of text and number and in unrecognized form. Thus along with the result output
embedded the key value and its generation used as cipher text to the receiver. The algorithm flow
is shown in the architecture diagram shown in fig 2.
Fig.2. Pictorial illust
Merging of DNA Cryptography and Cellular Automata
if these fields are combined, security provided by them will be enhanced. The proposed work in
dynamic in nature it contributes for dynamic behavior in the system which
increases the overall security. DNA cryptography provides a secure way to encrypt the data.
ackers less possible to crack the system.
cryptosystem it will lead to achieve properties of both fields. This algorithm comprises the
technique of both DNA cryptography and cellular automata and it is named as “A hybrid
51)” [18] [19].
Working of proposed system
Initially the plain text is taken from the user either in file or direct input. The input file is intially
the binary parts with the help of key on the n parts after xor operation. Convert the resultant value
f text and number and in unrecognized form. Thus along with the result output
is shown in the architecture diagram shown in fig 2.
Fig.2. Pictorial illustration of DNA based Cellular Automata System
42
nced. The proposed work in
dynamic in nature it contributes for dynamic behavior in the system which
tography provides a secure way to encrypt the data.
This algorithm comprises the
named as “A hybrid
The input file is intially
the n parts after xor operation. Convert the resultant value
f text and number and in unrecognized form. Thus along with the result output

43
3.2. Proposed Algorithm
The proposed algorithm is detailed below in the form of pseudocode.
Input:F[file of any type], A[Automata for Plain text to AGTC data generation], c1[code book1],
c2[code book2], c3[code book3], c4[code book4]
Output: File F with the AGTC coded DNA sequence
Algorithm : Encryption
Step 1: The File F is converted into DNA sequence:
F→ F[DNA(U)] ; // AGTC sequence file.
Step 2: In the File F[DNA] convert the Uracil into Thymine acid
F[DNA(U)] → F[DNA(T)];
Step 3:The DNA sequence is converted into binary sequence, The DNA output file
of
F[DNA(T)]→Binary[F[DNA(T)]]; // Binary file
Step 4: Split the Binary Sequence into equivalent parts:
Binary[F[DNA(T)]]→ {Binary[F1[DNA(T)]],
→ Binary[F2[DNA(T)]],
→ Binary[F3[DNA(T)]],
→ Binary[F4[DNA(T)]]}
Step 5: Perform XOR operation with help of keys K1, K2, K3, K4 for all Binary
Sequences F1, F2, F3, F4. // key generated and used.
X1=(Binary[F1[DNA(T)]]) (XOR)K1
X2=(Binary[F2[DNA(T)]]) (XOR)K2
X3=(Binary[F3[DNA(T)]]) (XOR) K3
X4=(Binary[F4[DNA(T)]]) (XOR) K4
Step 6: Combine the all output of the binary sequence into a single sequence.
X= X1+X2+X3+X4;
Step 7: Apply the rule 51 for the resultant binary sequence X.
Now the state of x in time t is X(t).
RULE 51 (00110011) → s(t+1) = NOT(s_currentCell(t))
Next State of X(t) is X(t+1).
X(t+1) = Rule 51(X(t));
After applying rule, resultant will be binary sequence. // Initial state t is
transformed into t+1 next state.

3.3.Implementation of Algorithm
The algorithm is implemented in three
the user is converted into DNA sequence by the codebook1 and codebook 2 given below in table
1 and table 2 respectively.
This code book 1 and 2 converts the plain text into DNA Sequence for per
operation of the algorithm.
Code Book 2
Then convert DNA sequence contains uracil into thymine component, the now converted DNA
sequence is converted into binary sequence by based on the codebook 3 given below.
By using the codebook 3 the DNA sequence is converted into binary sequence, for performing
step 4,5and 6. Where it converts into binary sequence and perform xor operation with key.
Code Book 3
Apply rule 51 of cellular automata [20] for the binary sequence, and convert the bi
into DNA code word using the table 3 and Send the AGTC data to Automata machine and with
the help of the code book 4 and automata machine convert the DNA sequence into DNA code
word.
Table 1: Code Book 1
Step 8: Convert the Binary sequence into DNA sequence
X(t+1)→ X[DNA(t+1)];
S= X[DNA(t+1)]
Step 9: In the Sequence S convert the Thymine(T) into
S→ S(T);
Step 10:Send the S(T)→ Automata Machine A[S(T)]; // DNA code word is
generated
Step 11: Repeat the above steps until reach the end of the file;
Step 12: The Encrypted Data is the output of the automata machine
A[S(T)]→ Encrypted Sequence;
Step 13: Append the Key in the output file at the end
Encrypted File[F]→ Encrypted Sequence + key;
This output file Contains the DNA code sequence.
End Algorithm
Implementation of Algorithm
The algorithm is implemented in three stages, initially the plain text which is taken as input from
This code book 1 and 2 converts the plain text into DNA Sequence for performing the step 1
3 the DNA sequence is converted into binary sequence, for performing
Apply rule 51 of cellular automata [20] for the binary sequence, and convert the binary sequence
Table 1: Code Book 1 - Plain text to DNA Sequence
Convert the Binary sequence into DNA sequence
→ X[DNA(t+1)];
In the Sequence S convert the Thymine(T) into Uracil(U) acid.
→ Automata Machine A[S(T)]; // DNA code word is
Repeat the above steps until reach the end of the file;
The Encrypted Data is the output of the automata machine
quence;
Append the Key in the output file at the end
→ Encrypted Sequence + key;
This output file Contains the DNA code sequence.
44
stages, initially the plain text which is taken as input from
forming the step 1
3 the DNA sequence is converted into binary sequence, for performing
nary sequence

2: Code Book 2 - Numbers and Special Characters to DNA Sequence
Table 3: Code Book 3 - DNA Sequence to Binary Sequence
45

Finally append the key with the DNA code word and send to receiver i.e. encrypted
file.
4. RESULTS AND DISCUSSION
The proposed algorithm is implemented using Java and MySql as back end. The sample screen
shot of the implementation is shown in Fig.3.
decrypted values are stored. The following are the computational parameters which have been
considered for evaluation for the proposed algorithm,
i) Encryption and decryption time consumed by the data (bits)
ii) Space utilization (MB)
iii) Throughput
Table 5: Comparing the data based on the
Types of
Algorithm
Computation time (s) of
data (bits)
8 16 32
RSA 50 52 54
AES 48 49 50
DES 49 49 51
Proposed
Method
50 50 51
The encryption time is nothing but the time taken by the data to be encrypted and then to be
stored in the database similarly the decryption time is the time taken to obtain the plain text from
Table 4: Code Book 4 - DNA code word
Finally append the key with the DNA code word and send to receiver i.e. encrypted
ISCUSSION
shot of the implementation is shown in Fig.3. The database is created and the encrypted and
The following are the computational parameters which have been
considered for evaluation for the proposed algorithm,
Encryption and decryption time consumed by the data (bits)
Table 5: Comparing the data based on the encryption time and storage
Computation time (s) of
data (bits)
Average
time
(ms)
Throughput
(ms)
Storage
(MB)
64 128 256
56 58 60 55 9.163 82.26
50 53 56 51 9.881 82.15
51 52 55 51.16 9.853 82.35
51 52 53 50.26 10.02 82.03
46
Finally append the key with the DNA code word and send to receiver i.e. encrypted text data or
The database is created and the encrypted and
The following are the computational parameters which have been
Storage
(MB)
82.26
82.15
82.35
82.03

the cipher text. The standard algorithms which are taken for t
DES algorithms. The parameters which are taken for comparison are average encryption and
decryption time, throughput and storage. The comparison of the well
proposed algorithm is given in table 5. As
the bits individually, so the values of the database was assigned as zero initially and after
aggregating all the set of the bits the final value was observed. To calculate the computation time
of data different block size in bits are taken into account, they are 8,16,32,64,128 and 256 bits.
The throughput can be obtained by dividing the average time with the total size of the data bit and
it is expressed by the formulae:
Where, η is throughput, D is data and T is the average time consumed by the data.
Encryption
The Plain text given by the user is encrypted using this algorithm effectively by applying all the
steps discussed in the algorithm and rule 51 is also applied to generate the encrypted data.
Result Analysis
The result based on the throughput obtained for the different algorithms during the encryption
process in shown in fig 4. It denotes that the DNA based Cellular Automata cryptographic
algorithm is having better performance than the other cryptographic algorithm.
the cipher text. The standard algorithms which are taken for the comparison are RSA, AES and
decryption time, throughput and storage. The comparison of the well-known algorithms with the
proposed algorithm is given in table 5. As the storage size shows slight deviation on considering
a different block size in bits are taken into account, they are 8,16,32,64,128 and 256 bits.
T
D
n
i
n
∑
=
= 0
η
is throughput, D is data and T is the average time consumed by the data.
in the algorithm and rule 51 is also applied to generate the encrypted data.
Fig.3. Encryption output
denotes that the DNA based Cellular Automata cryptographic
algorithm is having better performance than the other cryptographic algorithm.
47
he comparison are RSA, AES and
known algorithms with the
the storage size shows slight deviation on considering
a different block size in bits are taken into account, they are 8,16,32,64,128 and 256 bits.
in the algorithm and rule 51 is also applied to generate the encrypted data.
denotes that the DNA based Cellular Automata cryptographic

48
Fig.4. Comparison of algorithm based on throughput for encryption
Fig 5 illustrates the result based on the based on the throughput obtained for the different
algorithms during the decryption process. It denotes that the DNA based Cellular Automata
cryptographic algorithm is having better performance than the other cryptographic algorithm.
Fig.5. Comparison of algorithm based on throughput for decryption
From these results, it is easy to observe that the existing cryptographic algorithms still have a
disadvantage in the encryption process over the DNA cryptographic algorithms in terms time
consumption and serially in throughput. On the other hand, it is easy to observe that existing
cryptographic algorithm have a disadvantage in the decryption process over other algorithm in
terms of time consumption and storage. So based on the observed results it is found that the DNA
based Cellular Automata algorithm has good performance when compared with AES, DES and
RSA.
5. CONCLUSION
This paper explores the DNA cryptography along with the finite state machine. In the proposed
algorithm data is secured at 4 levels, using : XOR, conversion in DNA, DNA is converted into
binary based sequence, and it is subjected to rule 51 of cellular automata. Binary sequence is
again converted into DNA Sequence. By using this algorithm and mechanism the generated
cipher text is quite difficult to crack. The proposed algorithm is powerful in terms of security

49
features. The proposed algorithm is compared with other standard cryptographic algorithms
against various parameters. The proposed algorithm can be further customized such that it can
secure cloud database and other vulnerable system. As for now, the proposed algorithm can be
used only for the text encryption it can also be extended to image encryption.
REFERENCES
1. Ferguson, N.,Schnier, B. and KonhoT, “Cryptography Engineering: Design principles and Practical
applications”, 2010
2. Thiyagarajan P, Aghila G, Prasanna Venkatesan V, “Dynamic Pattern Based Image
Steganography”, Journal of Computing , Volume 3, Issue 2 February 2011, pp.1-9, ISSN 2151-
9617
3. Thiyagarajan P, Natarajan V, Aghila G, Prasanna Venkatesan V, Anitha R, “Pattern based 3D Image
Steganography”, Springer 3D Research Journal, Vol. 04, no.1, March 2013, pp.1-8, ISSN: 2092-6731
4. Thiyagarajan P, Aghila G, Prasanna Venkatesan V, “Stepping up Internet Banking Security
using Dynamic Pattern Based Image Steganography”, Springer (LNCS) in Communications in
Computer and Information Science Series(CCIS) , June 2011, ISSN: 1865:0929
5. U.Noorul Hussain, T.Chithralekha, “Review of DNA cryptology”, CiiT International Journal of
Networking and Communication Engineering, vol 3, No 13, October 2011, pg No 843-849.
6. Grasha Jacob, A. Murugan, “DNA based Cryptography: An Overview and Analysis”, ISSN: 2222-
4254, Int. J. Emerg. Sci., 3(1), 36-42, March 2013.
7. H. Z. Hsu and R. C. T. Lee, “DNA Based Encryption Methods”, The 23rd workshop on combinatorial
mathematics and computation theory.
8. Guangzhao Cui, Cuiling Li, Haobin Li, Xiaoguang Li “DNA Computing and Its Application to
Information Security Field”, 2009 fifth international conference on natural computation.
9. Anupriya Aggarwal, Praveen Kanth, "Secure Data Transmission Using DNA Encryption", Computer
Engineering and Intelligent Systems, Vol.5, No.7, 2014.
10. Sambhu Prasad Panda, MadhusmitaSahu, Umesh Prasad Rout, Surendra Kumar Nanda “Encryption
and Decryption algorithm using two dimensional cellular automata rules in Cryptography”,
International Journal of Communication Network & Security, Volume-1, Issue-1, 2011.
11. Rama R, BalaSuyambu J, Andrew Arokiaraj, ShanmugamSaravanan, “A Study of DES Algorithm
with cellular automata”, International Journal of Innovative Management, Information & Production
ISME International ⓒ2011 ISSN 2185-5439 Volume 3, Number 1,March 2012PP.10-16
12. Pratibha Sharma, NiranjanLal, ManojDiwakar “Text Security using 2D Cellular Automata Rules”,
Conference on Advances in Communication and Control Systems 2013 (CAC2S 2013).
13. RupaliSoni*, GopalPrajapati, Arif Khan, Deepak Kulhare, “Triple Stage DNA Cryptography Using
Sequential Machine”, International Journal of Advanced Research in Computer Science and Software
Engineering, Volume 3, Issue 8, August 2013 ISSN: 2277 128X.
14. M.Venugopal, Dr. E.G.Rajan, Dr. Sharma, “Security measures for Smart Devices through
Cryptography Using Cellular Automaton”, M.Venugopal et al, /(IJCSIT) International Journal of
Computer Science and Information Technologies, Vol. 4 (1) , 2013, 129 – 133.
15. PradiptaMajiand Chandrama Shaw, NiloyGanguly, Biplab K. Sikdar and P. Pal Chaudhuri, “Theory
and Application of Cellular Automata For Pattern Classification”, FundamentaInformaticae 58 (2003)
321–354 321,IOS Press.
16. Stephen Wolfram, “Random sequence generation by cellular automata”. Advances in Applied
Mathematics, 7:123,169, 1986.
17. Cellular automaton - Wikipedia, the free encyclopedia
url:http://en.wikipedia.org/wiki/Cellular_automaton retrieved on july 15, 2015
18. Cellular Automata rules lexicon - User DLLs - url:http://psoup.math.wisc.edu/mcell/rullex life.html
retrieved on Aug 26, 2015.
19. Cellular Automaton -- from Wolfram MathWorld
url:http://mathworld.wolfram.com/UniversalCellularAutomaton.html retrieved on aug 31, 2015.
20. Cellular Automata The 256 Rules (Stanford Encyclopedia of Philosophy)-
url:http://www.roguebasin.com/index.php retrieved on september 16, 2013.
21. S.Jeevidha et.al , “Analysis on DNA based Cryptography to Secure Data Transmission”, International
Journal of Computer Applications (0975 – 8887), Vol 29– No.8, September 2011.

A Novel DNA Encryption System Using Cellular Automata

More Related Content

Similar to A Novel DNA Encryption System Using Cellular Automata (20)

More from ClaraZara1 (20)

Recently uploaded (20)

A Novel DNA Encryption System Using Cellular Automata