You are on page 1of 7

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013)

Survey Report on Cryptography Based on Neural Network


Adel A. El-Zoghabi1, Amr H. Yassin2, Hany H. Hussien3
1
Head, Department of Information Technology, Institute of Graduate Studies and Research, Alexandria University
2
Lecturer, Electronics & Communication, Engineering Department, Alexandria Higher Institute of Engineering and Technology
3
PHD students, Department of Information Technology, Institute of Graduate Studies and Research, Alexandria University
Abstract Cryptography is the ability of changing
information into obvious unintelligibility in a way allowing a
secret method of un-mangling. The vital idea of cryptography
is the capability to send information between participants in a
way that prevents others from reading it. Much cryptography
methods are available which are based on number theory but
it has the disadvantage of requirement a large computational
power, complexity and time consumption. To overcome these
drawbacks, artificial neural networks (ANNs) have been
applied to solve many problems. The ANNs have many
characteristics such as learning, generalization, less data
requirement, fast computation, ease of implementation, and
software and hardware availability, which make it very
attractive for many applications. This paper provides a state-
of-the-art review on the use of artificial neural networks in
cryptography and studies their performance on
approximation problems related to cryptography. FIGURE 1: CRYPTOGRAPHY PUBLIC KEY COMPONENT [1]

Keywords Artificial neural network, chaotic map, Integrity: Assuring the receiver that the received
Cryptography, Decryption, Encryption, Key generation. message has not been altered in any way from the
original.
I. INTRODUCTION Non-repudiation: A mechanism to prove that the
Cryptography can be defined as the exchange of data sender really sent this message.
into a mingle code that can be deciphered and sent across a Authentication: The process of proving one's identity.
public or private network. Cryptography is the practice and Privacy/confidentiality: Ensuring that no one can read
study of hiding information. It is a critical part of secure the message except the intended receiver.
communication. Cryptograph not only protects data from A system that provides encryption and decryption is
robbery or alternation but can be used as well for user referred to as a cryptosystem and can be created through
authentication [1]. hardware components or program code in an application.
Cryptography has two core styles of encrypting data; The most cryptosystem algorithms are complex
symmetrical and asymmetrical. Symmetric encryptions use mathematical formulas that are applied in a specific
the same key for encryption and decryption process, and sequence to the plaintext. Most encryption methods use a
also can be defined as a secret-key, shared-key, and secret value called a key (usually a long string of bits),
private-key. Asymmetric cryptography uses different which works with the algorithm to encrypt and decrypt the
encryption keys for encryption and decryption process. In text [2].
this case an end user on a network, public or private, has a In the field of cryptography, one is interested in methods
pair of keys; one for encryption and one for decryption [1]. to transmit messages secretly between two parts. One (an
These keys can be identified as a public and a private key, opponent) who is able to listen to the communication
which can be shown in (Figure 1). should not be able to recover the secret message. Today a
Within the context of any application-to-application common secret key based on number theory could be
communication, there are some specific security created over a public channel accessible to any opponent
requirements, which including: but it might not be possible to calculate the discrete
logarithm of sufficiently large numbers.

456
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013)
DES, IDEA, RC5, CAST, BLOWFISH, 3DES, and RSA B. General regression neural networks
are the well-known and the mostly used (preferred) A memory-based network that provides estimates of
encryption and decryption systems. In general the cost of continuous variables and converges to the underlying
these systems is high and it requires more time for (linear or nonlinear) regression surface is described. The
computation and applications. general regression neural network (GRNN) is a one-pass
Artificial neural networks are parallel adaptive networks learning algorithm with a highly parallel structure. It is
consisting of simple nonlinear computing elements called shown that, even with sparse data in a multidimensional
neurons which are intended to abstract and model some of measurement space, the algorithm provides smooth
the functionality of the human nervous system in an transitions from one observed value to another. The
attempt to partially capture some of its computational algorithmic form can be used for any regression problem in
strengths. Neural networks are non-linear statistical data which an assumption of linearity is not justified.
modeling tools.
The development of ANNs comes from simulating C. Chaotic neural networks
intelligent tasks which are performed by human brain. They Chaotic neural networks offer greatly increase memory
are most widely used by soft computing techniques that capacity. Each memory is encoded by an Unstable Periodic
have the capability to capture and model complex Orbit (UPO) on the chaotic attractor. A chaotic attractor is
input/output relationships of any system. The advantages of a set of states in a system's state space with very special
ANNs are the ability to generalize results obtained from property that the set is an attracting set. So the system
known situations to unforeseen situations, the fast response starting with its initial condition in the appropriate basin,
time in operational phase, the high degree of structural eventually ends up in the set. The most important, once the
parallelism, reliability and efficiency. If a set of input- system is on the attractor nearby states diverge from each
output data pairs which belongs to a problem is available, other exponentially fast, however small amounts of noise
ANNs can learn and exhibit good performance. For these are amplified. According to a binary sequence generated
reasons, application of ANNs has emerged as a promising from a chaotic system, the biases and weights of neurons
area of research, since their adaptive behaviors have the are set.
potential of conveniently modeling strongly nonlinear
D. Multilayer neural networks (MLP)
characteristics.
The rest of the paper is organized as follows: Section 2 MLPs consist of multiple layers of computational units,
describes the main ANNs found in the studied literature. interconnected in a feed-forward way. MLPs use a variety
Section 3 presents the survey on various neural network of learning techniques, the most popular being back-
techniques and related work to cryptography. Finally, propagation, where the output values are compared with the
Section 4 shows summary points and Section 5 discusses correct answer to compute the value of some predefined
the conclusions error-function. The error is then fed back through the
network. Using this information, the algorithm adjusts the
II. TYPES OF ARTIFICIAL NEURAL NETWORK (ANNS) weights of each connection in order to reduce the value of
the error function by some small amount.
After analyzing the aforementioned articles, various
ANN topologies were found. The most frequent ones are E. Neural cryptography
described below: Is a branch of cryptography dedicated to analyzing the
A. Recurrent Neural Networks (RNN) application of stochastic algorithms, especially neural
network algorithms, for use in encryption and
Is a network which neurons send feedback signals to cryptanalysis. Neural Networks are well known for their
each other, such as the Hopfield network, Elman and ability to selectively explore the solution space of a given
Jordans network, the Long Short Time Memory RNN and
problem. This feature finds a natural niche of application in
bidirectional networks. This enables the modeling of
the field of cryptanalysis.
dynamic behaviors with the drawback of more memory
consumption if compared to direct networks.

457
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013)
At the same time, Neural Networks offer a new approach The both partners used their neural networks as input for
to attack ciphering algorithms based on the principle that the logistic maps which generated the output bits to be
any function could be reproduced by a neural network, learned, by mutually learning. The two neural networks
which is a powerful proven computational tool that can be approach each other and generated a matching signal to the
used to find the inverse-function of any cryptographic chaotic maps. The chaotic synchronization applied in the
algorithm. The ideas of mutual learning, self learning, and neural cryptography enhanced the cryptography systems
stochastic behavior of neural networks and similar and improved the security [4].
algorithms can be used for different aspects of
A New Security on Neural Cryptography with Queries,
cryptography, like public-key cryptography, solving the
2010
key distribution problem using neural network mutual
synchronization, hashing or generation of pseudo-random N. Prabakaran proposed a secret key using neural
numbers. cryptography, based on synchronization of Tree Parity
Machines (TPMs) by mutual learning. The system has two
III. LITERATURE REVIEW identical dynamical systems, which starting from different
initial conditions and synchronized by a common input
A recent survey of the literature indicates that there has values which are coupled to the two systems. The networks
been an increasing interest in the application of different received a common input vector after calculating their
classes of neural networks to problems related to outputs and updated their weight vectors according to the
cryptography in the past few years. Recent works have match between their mutual outputs in every time step. The
examined the use of neural networks in cryptosystems [3- input or output relations are not exchanged through a public
23]. This can be categorized into three sections: channel until their weight vectors are matching and can be
A. Synchronization neural networks used as a secret key for encryption and decryption of secret
messages. The weight vectors of the two neural networks
Neural Cryptography, 2003 begin with random numbers, which are generated by
Wolfgang Kinzel proposed a secret key over a public Pseudo-Random Number Generators (PRNGs). The
channel using artificial neural networks. The artificial proposed model fixed the security against numerical attacks
neural network contains of two multi-layer neural networks [5].
trained on their mutual output bits and able to synchronize.
The two networks starting from random initial weights and Design of an efficient neural key generation, 2011
learning from each other with two multilayer networks R. M. Jogdand proposed a common secret key generated
relax to the state with time dependent identical synaptic based on neural networks. The neural cryptography has
weights. The partners didnt exchange any information two communication networks that received an identical
over a secret channel before their communication. input vector, generated an output bit and are trained based
Synchronization of neural networks can be considered as on the output bit. The network model initials the weight
the key generation in cryptography. The common identical randomly and the input object is generated by another
weights of the two partners can be used as a key for source and the outputs bit are generated finally and
encryption. The neural cryptography is the first algorithm exchange between patterns. The weight may be modified if
for key generation over public channels which are not the outputs of both partners are matched. The modified
based on number theory. Experimental result shows that the weight after synchronize act as a secret key for the
model is fast, simple, and secures [3]. encryption and decryption process. Simulation results show
that the cryptosystem based on ANNs is secure [6].
Synchronization of neural networks by mutual learning and
its application to cryptography, 2004 Cryptography in structure adaptable digital neural
Einat Klein presented a secured cryptography secret-key networks, 2012
based on neural network in a public channel. The proposed Pratap Singh generated a secret key over a public
model has two neural networks that are trained on their channel based on neural network. The model has a neural
alternate output synchronized to an equal time dependent network machine contains of two partners started with
weight vector through a chaos synchronization system that initial weights and different initial conditions which
starting from different initial conditions. The system synchronized by a common external signal and received a
combined the neural networks with the logistic chaotic common random input vector and learned their mutual
map. output bits.
458
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013)
The synaptic weights are used as a secret key over a Power consumption using artificial neural network in the
public channel. Simulation results show that the model are field of cryptography, 2012
secure and efficiency [7]. Rajender S. presented a triple key chaotic neural
Neural Cryptography for Secret Key Exchange and network for cryptography. The chaotic neural network are
Encryption with AES, 2013 used for generating a binary sequences for masking
plaintext .The plaintext is masked by switching of chaotic
Ajit Singh presented synchronization neural key-
neural network and permutation of generated binary
exchange algorithm for cryptography. The model has
sequences. The triple key chaotic neural networks are
multi-layer feed-forward neural network which have two
contained of 20 hexadecimal characters that entered as
tree parity machine (TPM) that synchronized with a
session key with some extraction and manipulations
random initial weight act as common secret key for the
process to get intermediate key, which combined with the
encryption and decryption process. The weight can be
initial value and control parameters to generate the triple
updated according to the learning rule only if the output
key chaotic sequence. Experimental results show that the
values of the two machines are equal. Throughout the
model is highly secure, but with a little concern about time
synchronization process, only the input vectors and the
consumption [11].
output vectors are transmitted over the public channel.
Experimental results show that the model are efficiency A triple-key chaotic neural network for cryptography in
and secure through increasing the common key size [8]. image processing, 2012
B. Chaotic Neural network Shweta B. presented a triple key chaotic neural network
for image cryptography. The triple parameters are used to
Cryptography based on delayed chaotic neural networks, perform the various operations on image so as to scramble
2006 the data in particular way which look like random but
Wenwu Yu proposed an encryption techniques based on actually it is in particular sequence. The triple key contains
the chaotic hopfield neural networks with time varying a hexadecimal key that extraction and manipulations to
delay. The chaotic neural network is used for generating achieve the intermediate key which combined with initial
binary sequences for masking the plaintext. The binary and control parameters to generate chaotic sequence.
value of the binary sequence chooses the chaotic logistic Experimental results shows that algorithm successfully
map randomly, that used for generated the binary perform the cryptography and can be applied on different
sequences. The plaintext is masked by switching of the colour image size [12].
chaotic neural network maps and permutation of generated
An Empirical Investigation of Using ANN Based N-State
binary sequences. Simulation results show that the
Sequential Machine and Chaotic Neural Network in the
proposed chaotic cryptography is more functional in the
Field of Cryptography, 2012
secure transmission of large multi-media files over public
data communication network [9]. Nitin Shukla proposed two artificial neural networks for
cryptography. The First network is neural network based n-
Cryptanalysis of a cryptographic scheme based on delayed state sequential machine and other one is chaotic neural
chaotic neural networks, 2009 network. The first network generated a finite state
Jiyun Yang analysed the proposed model of Wenwu Yu sequential machine using simple recurrent neural network
[9]. It was difficult to obtain the key of Yu et al.s based on back propagation training algorithm. The starting
cryptosystem through classical attacks because of large key state of the n-state sequential machine can be used as a key
space. However, as the same key stream is used in every for encryption and decryption process. The second network
encryption process, it can be easily obtained by the chosen divided the message into blocks and identified the initial
plaintext attack using two pairs of plaintext and cipher- value, and the control parameter, then generated the chaotic
texts only. Simulation results show that the proposed sequence. The weights and biases of the neural network are
chaotic cryptography is insecure [10]. determined based on the chaotic sequence, and act as a key
for encryption and decryption process. Experimental results
show that the two networks are secure, without any results
about efficiency [13].

459
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013)
Encryption of mpeg-2 video signal based on chaotic neural Each layer consists of the following, an input layer
network, 2012 consisting of 3 nodes, which represent the n-bit blocks, a
Tarip A. proposed adapted system model containing pattern layer of 8 nodes, and an output layer of 8 nodes,
MPEG-2 for compression and a chaotic neural network which used to identify the decrypted output message. The
(CNN) for cryptography. The algorithm supports quality simulation results have shown a very good result, with
and bit rate control that is required by many video relatively better performance than the traditional encryption
transmission applications, which is considered a new methods, but without any knowledge about the security of
hopeful field related to video protection and compression. the proposed model [17].
The logistics map is used with neural network to produce a A Back propagation Neural Network for Computer
combination of CNN, based on a binary sequence Network Security, 2006
generated from the logistic map, the biases and weights of
Khalil Shihab provided a new asymmetric encryption
neurons are modified and considerer to be the secret key.
mechanism based on artificial neural networks. The
The required time for CNN encryption and decryption can
decryption scheme and the public key system creation
be reduced by increased the size of neural network from 64
process are based on multi-layer neural network. The
to 256 weights. It has been shown from analysis results that
neural network is trained by back-propagation learning
the proposed algorithm has high security with low cost, and
algorithm, while the encryption scheme and the private key
also supports quality and bit rate control [14].
creation process are based on Boolean algebra. The
Use of Artificial Neural Network in the Field of Security, cryptography schemes are not based on number theoretic
2013 functions and have a small time and memory complexities.
Navita A. proposed two artificial neural networks for Experimental results show that the security and efficiency
cryptography. This model is identical to the model of the proposed model is better than or equal to the
presented by Nitin Shukla[13] . Experimental results show traditional methods [18].
that the two networks are secure, without any results about Neural Solutions for Information Security, 2007
efficiency [15].
Seref S. presented a new data security approach over
A Hybrid Model for Secure Data Transfer in Audio Signals electronic communication based on artificial neural
using HCNN and DD DWT, 2013 networks (ANNs). The Neural network model converted
Geetha vani proposed a new combined model for the plain message into binary form and applied the neural
cryptography and steganography. The model using the network model on it to get different cipher message
Hopfield Chaotic Neural Network (HCNN) for the sequence. The key source of the neural network encryption
cryptography which uses the chaotic trajectories of two module are the number of neurons in the input, hidden and
neurons to produce main binary sequences for encryption output layers, the number of hidden layer, the weights, the
the plain-text. The model used also the Double Density biases and the type of transfer function used in layers,
Discrete Wavelet Transform (DD DWT) to embed the which transferred from sender to receiver only once in a
cipher-text into the audio cover. Experimental results show secure channel. After receiving the source key from sender,
that the model is efficiency and secure against the most the neural network decryption module is reconstructed at
knows attacks [16]. receiver and the message can be decrypted after received
via insecure channel. Extra security can be attained by
C. Multi-layer Neural networks applied extra module which inserts random binary numbers
Data security based on neural networks, 2005 into the encrypted message line within a sequence and
provided more complexity to the encrypted messages and
Khaled M. presented a cryptography model based on the
also increased the security. Simulation results show that the
general regression neural network (GRNN). The proposed
cryptosystem based on ANNs is very efficiency, provides
GRNN has three layers; each layer consists of a number of
more security, and can be applied for real-time application
neurons, depending on the process type. The encryption
[19].
process divided the input message into 3-bit data sets, and
produced 8-bit after the encryption process.

460
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013)
Artificial neural network based chaotic generator for The generation process of the key using back
cryptology, 2010 propagation neural network consists of three phases. The
Ilker D. proposed a chaotic cryptosystems based on first phase is the feed-forward, the second phase is a back-
artificial neural network and chaotic generator propagation of the associated error and the third phase is
synchronization. The ANN model generated the chaotic related with weights adjustments. Simulation results show
dynamics by the numerical solution of Chuas circuit. The that the cryptosystem based on ANNs is very efficiency,
proposed model have three initial conditions and time and can be applied on hardware application [22].
variable as input with two hidden layers and three chaotic Cryptography based on neural network, 2012
dynamics as output. Many simulations are done on the
Eva Volna proposed multi-layer neural networks in
number of neurons on hidden layer to find the best ANN
cryptography. The multilayer neural networks modified by
structure obtained chaotic dynamics. The chaotic dynamics
back-propagation. The planned model converted the input
act as the key for encryption and decryption process. The
message intro ASCII code then gets the sequence of bit for
ANN model does not have any synchronization problem.
each code which divided into 6 bit blocks are used as input
The difference between the chaotic dynamics can be
for the encryption process. The cipher key is the neural
considered as an advantage of the ANN based chaotic
network structure contained input layer, hidden layer,
generator. The major weaknesses of analogue circuit and
output layer, and updated weights. Experimental results
the numerical solution of chaotic circuit are eliminated with
show that the system is secure [23].
the proposed model. Simulation results show that the model
are efficiency, secure, and can be applying on real time
application [20]. IV. SUMMARY
In this paper, many of the important ANN techniques
AES Cryptosystem Development Using Neural Networks, have been presented and analyzed. These techniques are
2011 based on:
Siddeeq. Y presented a new modification of the
Advanced Encryption Standard (AES) to be immune
The Tree Parity Machines (TPM) was used to
generate a secret key over the public channel on the
against some attacks using nonlinear neural network. The
output at each partner.
neural network model performs cryptography processes via
a symmetric key cipher that used as the initial weights for The neural network with chaotic logistic map was
the neural network which trained to its final weight fast and used for cryptography by which both partners use the
low cost algorithm. The objective from the network has neural network as input for the logistic map, that
been selected to equivalent the output of the AES that have generate the output bits to be learned.
an efficient and recommended security. Simulation results The General Regression Neural Network (GRNN)
show the proximity of the results accomplished by the was used for encryption and decryption process based
proposed NN-based AES cryptosystem with that of the on three layers, where the input data divided into 3
normal AES [21]. bits and 8 bits as output.
The training back-propagation neural network was act
Implementation of neural - cryptographic system using as public key, while Boolean algebra act as private
FPGA, 2011 key.
Karam M. presented a stream cipher system based on The chaotic hopfield neural network with time
pseudo Random Number Generator (PRNG) through using varying delay was used to generate binary sequence
artificial Neural Networks (ANN). The PRNG model has a for making plaintext, which considered as a random
high statistical randomness properties for key sequence switching function for chaotic map.
using ANN. The proposed neural pseudo random number The neural network based on chaotic generator was
generator consists of two stages; the first stage is used for generate chaotic dynamic act as a shard key.
generating a long sequence of patterns from perfect The initial weight value of the neural network was
equation and initial value. So these patterns possess the used after training as symmetric key.
randomness and unpredictable properties. The total number The Pseudo Random Number Generator (PRN) based
of equations and initial values depend on the number of bits on neural network was used in stream cipher as a key
that represented the initial value, the second stage is an sequence generator.
artificial neural network (ANN) that gets the outputs of the
previous stage and set it as input to the NN.
461
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013)
The chaotic neural network was used to generate [9] Wenwu Yu, Jinde Cao, Cryptography based on delayed chaotic
neural networks, Physics Letters A, Vol. 356, (4) Elsevier, 333
chaotic sequence act as a triple key (combined of 338, 2006.
initial condition and control parameters) for [10] Jiyun Yang, Xiaofeng Liao, Wenwu Yu, Kwok-wo Wong, Jun Wei,
cryptography. Cryptanalysis of a cryptographic scheme based on delayed chaotic
The Layer Recurrent Neural Network (LRNN) was neural networks, Chaos, Solitons& Fractals, Vol. 40, Issue.2, 821-
used to generate pseudo random number based on 825, 2009.
weight matrix obtained from layer weight of the [11] Rajender Singh, Rahul Misra, Abhishek Chaudhary, Power
consumption using artificial neural network in the field of
LRNN. cryptography, Journal of information, Knowledge and research in
computer Engineering, Vol.2, Issue.2, 443- 446, 2012.
V. CONCLUSION [12] Shweta B. Suryawanshi, Devesh D. Nawgaje, A triple-key Chaotic
neural network for cryptography in image processing, International
In this paper, we summarize some recent researches Journal of Engineering Sciences & Emerging Technologies, Vol. 2,
about the application of neural network in the field of Issue. 1, 46-50, 2012.
cryptography. The designed NN-based cryptosystem is a [13] Nitin Shukla, Abhinav Tiwari, An Empirical Investigation of Using
good idea of building very complicated cryptosystem, ANN Based N-State Sequential Machine and Chaotic Neural
where the crypto analyst or the cracker not just need the Network in the Field of Cryptography, Global Journal of Computer
Science and Technology Neural & Artificial Intelligence, Vol. 12,
topology of the NN and the key to crack the system, but Issue.10,No. 1,17-26, 2012.
also need to know the number of adaptive iterations and the [14] Tariq A. fadil, Shahrul N. yaakob, Badlishah ahmad, abid yahya,
final weights for the encryption and decryption systems. Encryption of mpeg-2 video signal based on chaotic neural
Applying higher numbers of plain-text/ cipher-text to the network, Journal of Engineering and Technology, Vol. 3, 35-42,
NN-based cryptosystem so as to make the error rate as 2012.
minimum as possible. [15] Navita Agarwal, Prachi Agarwal, Use of Artificial Neural Network
in the Field of Security, MIT International Journal of Computer
Lastly, according to the findings of this work, the trend Science & Information Technology, Vol. 3, No. 1, 4244, 2013.
for the years to come regarding the use of ANNs for
[16] B. Geetha vani, E. V. Prasad, A Hybrid Model for Secure Data
cryptography tasks will be focused mainly on TPM, CNNs, Transfer in Audio Signals using HCNN and DD DWT, (IJACSA)
and LRNN. International Journal of Advanced Computer Science and
Applications, Vol. 4, No.7, 202-208, 2013.
REFERENCES [17] Khaled M. G. Noaman, Hamid Abdullah Jalab, Data security based
[1] William Stallings, Cryptography and Network Security: Principles on neural networks, Task Quarterly 9, No. 4, 409414, 2005.
and Practice, (5th Edition), Prentice Hall, 2010. [18] Khalil Shihab, A Back propagation Neural Network for Computer
[2] T. P.Wasnik , Vishal S. Patil , Sushant A. Patinge , Sachin R. Dave , Network Security, Journal of Computer Science 2 (9): 710-715,
Gaurav J. Sayasikamal, Cryptography as an instrument to network 2006.
security, International Journal of Application or Innovation in [19] Seref S., Necla. , Neural Solutions for Information Security,
Engineering & Management (IJAIEM), Vol. 2, Issue 3, 72-80, 2013. Politeknik Dergisi, Journal of Polytechnic, Vol. 10, No. 1, 21-25,
[3] Wolfgang Kinzel, IdoKanter, Neural Cryptography, Proceedings 2007.
TH2002 Supplement, Vol. 4, 147 153, 2003. [20] Ilker Dalkiran, Kenan Danis, Artificial neural network based
[4] Einat Klein, Rachel Mislovaty, IdoKanter, Andreas Ruttor, chaotic generator for cryptology, Turk J Elec Eng& Comp Sci,
Wolfgang Kinzel, Synchronization of neural networks by mutual Vol.18, No.2, 255- 240, 2010.
learning and its application to cryptography, In proceeding of: [21] Siddeeq. Y. Ameen, Ali H. Mahdi, AES Cryptosystem
Advances in Neural Information Processing Systems 17 ,Neural Development Using Neural Networks, International Journal of
Information Processing Systems NIPS, 2004. Computer and Electrical Engineering, Vol. 3, No. 2, 315- 318, 2011.
[5] N. Prabakaran, P. Vivekanandan, A New Security on Neural [22] Karam M. Z. Othman, Mohammed H. AL Jammas, Implementation
Cryptography with Queries, Int. J. of Advanced Networking and of neural - cryptographic system using FPGA, Journal of
Applications, Vol. 2, Issue. 1, 437-444, 2010. Engineering Science and Technology, Vol. 6, No. 4, 411 428,
[6] R. M. Jogdand, Sahana S. Bisalapur, Design of an efficient neural 2011.
key generation, International Journal of Artificial Intelligence & [23] Eva Volna, Martin Kotyrba, Vaclav Kocian, Michal Janosek,
Applications (IJAIA), Vol.2, No.1, 60- 69, 2011. Cryptography based on neural network, Proceedings 26th
[7] Pratap Singh, Harvir Singh, Cryptography in structure adaptable European Conference on Modelling and Simulation, 2012.
digital neural networks", National monthly refereed journal of
research in science & technology, Vol.1, Issue.12, 35-44, 2012.
[8] Ajit Singh, Aartinandal, Neural Cryptography for Secret Key
Exchange and Encryption with AES, International Journal of
Advanced Research in Computer Science and Software Engineering,
Vol. 3, Issue.5, 376- 381, 2013.

462

You might also like