Survey Report on Cryptography Based on Neural Network - IJETAE

9 downloads 28366 Views 318KB Size Report
Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013). 456 .... ANNs can learn and exhibit good performance. For these ..... Artificial neural network based chaotic generator for.
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 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

2

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 stateof-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]

 Integrity: Assuring the receiver that the received message has not been altered in any way from the original.  Non-repudiation: A mechanism to prove that the sender really sent this message.  Authentication: The process of proving one's identity.  Privacy/confidentiality: Ensuring that no one can read the message except the intended receiver. A system that provides encryption and decryption is referred to as a cryptosystem and can be created through hardware components or program code in an application. The most cryptosystem algorithms are complex mathematical formulas that are applied in a specific sequence to the plaintext. Most encryption methods use a secret value called a key (usually a long string of bits), which works with the algorithm to encrypt and decrypt the text [2]. In the field of cryptography, one is interested in methods to transmit messages secretly between two parts. One (an opponent) who is able to listen to the communication should not be able to recover the secret message. Today a common secret key based on number theory could be created over a public channel accessible to any opponent but it might not be possible to calculate the discrete logarithm of sufficiently large numbers.

Keywords— Artificial neural network, chaotic map, Cryptography, Decryption, Encryption, Key generation.

I. INTRODUCTION Cryptography can be defined as the exchange of data into a mingle code that can be deciphered and sent across a public or private network. Cryptography is the practice and study of hiding information. It is a critical part of secure communication. Cryptograph not only protects data from robbery or alternation but can be used as well for user authentication [1]. Cryptography has two core styles of encrypting data; symmetrical and asymmetrical. Symmetric encryptions use the same key for encryption and decryption process, and also can be defined as a secret-key, shared-key, and private-key. Asymmetric cryptography uses different encryption keys for encryption and decryption process. In this case an end user on a network, public or private, has a pair of keys; one for encryption and one for decryption [1]. These keys can be identified as a public and a private key, which can be shown in (Figure 1). Within the context of any application-to-application communication, there are some specific security requirements, which including:

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 are the well-known and the mostly used (preferred) encryption and decryption systems. In general the cost of these systems is high and it requires more time for computation and applications. Artificial neural networks are parallel adaptive networks consisting of simple nonlinear computing elements called neurons which are intended to abstract and model some of the functionality of the human nervous system in an attempt to partially capture some of its computational strengths. Neural networks are non-linear statistical data modeling tools. The development of ANNs comes from simulating intelligent tasks which are performed by human brain. They are most widely used by soft computing techniques that have the capability to capture and model complex input/output relationships of any system. The advantages of ANNs are the ability to generalize results obtained from known situations to unforeseen situations, the fast response time in operational phase, the high degree of structural parallelism, reliability and efficiency. If a set of inputoutput data pairs which belongs to a problem is available, ANNs can learn and exhibit good performance. For these reasons, application of ANNs has emerged as a promising area of research, since their adaptive behaviors have the potential of conveniently modeling strongly nonlinear characteristics. The rest of the paper is organized as follows: Section 2 describes the main ANNs found in the studied literature. Section 3 presents the survey on various neural network techniques and related work to cryptography. Finally, Section 4 shows summary points and Section 5 discusses the conclusions

B. General regression neural networks A memory-based network that provides estimates of continuous variables and converges to the underlying (linear or nonlinear) regression surface is described. The general regression neural network (GRNN) is a one-pass learning algorithm with a highly parallel structure. It is shown that, even with sparse data in a multidimensional measurement space, the algorithm provides smooth transitions from one observed value to another. The algorithmic form can be used for any regression problem in which an assumption of linearity is not justified. C. Chaotic neural networks Chaotic neural networks offer greatly increase memory capacity. Each memory is encoded by an Unstable Periodic Orbit (UPO) on the chaotic attractor. A chaotic attractor is a set of states in a system's state space with very special property that the set is an attracting set. So the system starting with its initial condition in the appropriate basin, eventually ends up in the set. The most important, once the system is on the attractor nearby states diverge from each other exponentially fast, however small amounts of noise are amplified. According to a binary sequence generated from a chaotic system, the biases and weights of neurons are set. D. Multilayer neural networks (MLP) MLPs consist of multiple layers of computational units, interconnected in a feed-forward way. MLPs use a variety of learning techniques, the most popular being backpropagation, where the output values are compared with the correct answer to compute the value of some predefined error-function. The error is then fed back through the network. Using this information, the algorithm adjusts the weights of each connection in order to reduce the value of the error function by some small amount.

II. TYPES OF ARTIFICIAL NEURAL NETWORK (ANNS) After analyzing the aforementioned articles, various ANN topologies were found. The most frequent ones are described below:

E. Neural cryptography Is a branch of cryptography dedicated to analyzing the application of stochastic algorithms, especially neural network algorithms, for use in encryption and cryptanalysis. Neural Networks are well known for their ability to selectively explore the solution space of a given problem. This feature finds a natural niche of application in the field of cryptanalysis.

A. Recurrent Neural Networks (RNN) Is a network which neurons send feedback signals to each other, such as the Hopfield network, Elman and Jordan’s network, the Long Short Time Memory RNN and bidirectional networks. This enables the modeling of 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 to attack ciphering algorithms based on the principle that any function could be reproduced by a neural network, which is a powerful proven computational tool that can be used to find the inverse-function of any cryptographic algorithm. The ideas of mutual learning, self learning, and stochastic behavior of neural networks and similar algorithms can be used for different aspects of cryptography, like public-key cryptography, solving the key distribution problem using neural network mutual synchronization, hashing or generation of pseudo-random numbers.

The both partners used their neural networks as input for the logistic maps which generated the output bits to be learned, by mutually learning. The two neural networks approach each other and generated a matching signal to the chaotic maps. The chaotic synchronization applied in the neural cryptography enhanced the cryptography systems and improved the security [4]. A New Security on Neural Cryptography with Queries, 2010 N. Prabakaran proposed a secret key using neural cryptography, based on synchronization of Tree Parity Machines (TPMs) by mutual learning. The system has two identical dynamical systems, which starting from different initial conditions and synchronized by a common input values which are coupled to the two systems. The networks received a common input vector after calculating their outputs and updated their weight vectors according to the match between their mutual outputs in every time step. The input or output relations are not exchanged through a public channel until their weight vectors are matching and can be used as a secret key for encryption and decryption of secret messages. The weight vectors of the two neural networks begin with random numbers, which are generated by Pseudo-Random Number Generators (PRNGs). The proposed model fixed the security against numerical attacks [5].

III. LITERATURE REVIEW A recent survey of the literature indicates that there has been an increasing interest in the application of different classes of neural networks to problems related to cryptography in the past few years. Recent works have examined the use of neural networks in cryptosystems [323]. This can be categorized into three sections: A. Synchronization neural networks Neural Cryptography, 2003 Wolfgang Kinzel proposed a secret key over a public channel using artificial neural networks. The artificial neural network contains of two multi-layer neural networks trained on their mutual output bits and able to synchronize. The two networks starting from random initial weights and learning from each other with two multilayer networks relax to the state with time dependent identical synaptic weights. The partners didn’t exchange any information over a secret channel before their communication. Synchronization of neural networks can be considered as the key generation in cryptography. The common identical weights of the two partners can be used as a key for encryption. The neural cryptography is the first algorithm for key generation over public channels which are not based on number theory. Experimental result shows that the model is fast, simple, and secures [3].

Design of an efficient neural key generation, 2011 R. M. Jogdand proposed a common secret key generated based on neural networks. The neural cryptography has two communication networks that received an identical input vector, generated an output bit and are trained based on the output bit. The network model initials the weight randomly and the input object is generated by another source and the outputs bit are generated finally and exchange between patterns. The weight may be modified if the outputs of both partners are matched. The modified weight after synchronize act as a secret key for the 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 Einat Klein presented a secured cryptography secret-key based on neural network in a public channel. The proposed model has two neural networks that are trained on their alternate output synchronized to an equal time dependent weight vector through a chaos synchronization system that starting from different initial conditions. The system combined the neural networks with the logistic chaotic map.

Cryptography in structure adaptable digital neural networks, 2012 Pratap Singh generated a secret key over a public channel based on neural network. The model has a neural network machine contains of two partners started with initial weights and different initial conditions which synchronized by a common external signal and received a common random input vector and learned their mutual 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 public channel. Simulation results show that the model are secure and efficiency [7].

Power consumption using artificial neural network in the field of cryptography, 2012 Rajender S. presented a triple key chaotic neural network for cryptography. The chaotic neural network are used for generating a binary sequences for masking plaintext .The plaintext is masked by switching of chaotic neural network and permutation of generated binary sequences. The triple key chaotic neural networks are contained of 20 hexadecimal characters that entered as session key with some extraction and manipulations process to get intermediate key, which combined with the initial value and control parameters to generate the triple key chaotic sequence. Experimental results show that the model is highly secure, but with a little concern about time consumption [11].

Neural Cryptography for Secret Key Exchange and Encryption with AES, 2013 Ajit Singh presented synchronization neural keyexchange algorithm for cryptography. The model has multi-layer feed-forward neural network which have two tree parity machine (TPM) that synchronized with a random initial weight act as common secret key for the encryption and decryption process. The weight can be updated according to the learning rule only if the output values of the two machines are equal. Throughout the synchronization process, only the input vectors and the output vectors are transmitted over the public channel. Experimental results show that the model are efficiency and secure through increasing the common key size [8].

A triple-key chaotic neural network for cryptography in image processing, 2012 Shweta B. presented a triple key chaotic neural network for image cryptography. The triple parameters are used to perform the various operations on image so as to scramble the data in particular way which look like random but actually it is in particular sequence. The triple key contains a hexadecimal key that extraction and manipulations to achieve the intermediate key which combined with initial and control parameters to generate chaotic sequence. Experimental results shows that algorithm successfully perform the cryptography and can be applied on different colour image size [12].

B. Chaotic Neural network Cryptography based on delayed chaotic neural networks, 2006 Wenwu Yu proposed an encryption techniques based on the chaotic hopfield neural networks with time varying delay. The chaotic neural network is used for generating binary sequences for masking the plaintext. The binary value of the binary sequence chooses the chaotic logistic map randomly, that used for generated the binary sequences. The plaintext is masked by switching of the chaotic neural network maps and permutation of generated binary sequences. Simulation results show that the proposed chaotic cryptography is more functional in the secure transmission of large multi-media files over public data communication network [9].

An Empirical Investigation of Using ANN Based N-State Sequential Machine and Chaotic Neural Network in the Field of Cryptography, 2012 Nitin Shukla proposed two artificial neural networks for cryptography. The First network is neural network based nstate sequential machine and other one is chaotic neural network. The first network generated a finite state sequential machine using simple recurrent neural network based on back propagation training algorithm. The starting state of the n-state sequential machine can be used as a key for encryption and decryption process. The second network divided the message into blocks and identified the initial value, and the control parameter, then generated the chaotic sequence. The weights and biases of the neural network are 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].

Cryptanalysis of a cryptographic scheme based on delayed chaotic neural networks, 2009 Jiyun Yang analysed the proposed model of Wenwu Yu [9]. It was difficult to obtain the key of Yu et al.’s cryptosystem through classical attacks because of large key space. However, as the same key stream is used in every encryption process, it can be easily obtained by the chosen plaintext attack using two pairs of plaintext and ciphertexts only. Simulation results show that the proposed chaotic cryptography is insecure [10].

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 network, 2012 Tarip A. proposed adapted system model containing MPEG-2 for compression and a chaotic neural network (CNN) for cryptography. The algorithm supports quality and bit rate control that is required by many video transmission applications, which is considered a new hopeful field related to video protection and compression. The logistics map is used with neural network to produce a combination of CNN, based on a binary sequence generated from the logistic map, the biases and weights of neurons are modified and considerer to be the secret key. The required time for CNN encryption and decryption can be reduced by increased the size of neural network from 64 to 256 weights. It has been shown from analysis results that the proposed algorithm has high security with low cost, and also supports quality and bit rate control [14].

Each layer consists of the following, an input layer consisting of 3 nodes, which represent the n-bit blocks, a pattern layer of 8 nodes, and an output layer of 8 nodes, which used to identify the decrypted output message. The simulation results have shown a very good result, with relatively better performance than the traditional encryption methods, but without any knowledge about the security of the proposed model [17]. A Back propagation Neural Network for Computer Network Security, 2006 Khalil Shihab provided a new asymmetric encryption mechanism based on artificial neural networks. The decryption scheme and the public key system creation process are based on multi-layer neural network. The neural network is trained by back-propagation learning algorithm, while the encryption scheme and the private key creation process are based on Boolean algebra. The cryptography schemes are not based on number theoretic functions and have a small time and memory complexities. Experimental results show that the security and efficiency of the proposed model is better than or equal to the traditional methods [18].

Use of Artificial Neural Network in the Field of Security, 2013 Navita A. proposed two artificial neural networks for cryptography. This model is identical to the model presented by Nitin Shukla[13] . Experimental results show that the two networks are secure, without any results about efficiency [15].

Neural Solutions for Information Security, 2007 Seref S. presented a new data security approach over electronic communication based on artificial neural networks (ANNs). The Neural network model converted the plain message into binary form and applied the neural network model on it to get different cipher message sequence. The key source of the neural network encryption module are the number of neurons in the input, hidden and output layers, the number of hidden layer, the weights, the biases and the type of transfer function used in layers, which transferred from sender to receiver only once in a secure channel. After receiving the source key from sender, the neural network decryption module is reconstructed at receiver and the message can be decrypted after received via insecure channel. Extra security can be attained by applied extra module which inserts random binary numbers into the encrypted message line within a sequence and provided more complexity to the encrypted messages and also increased the security. Simulation results show that the cryptosystem based on ANNs is very efficiency, provides more security, and can be applied for real-time application [19].

A Hybrid Model for Secure Data Transfer in Audio Signals using HCNN and DD DWT, 2013 Geetha vani proposed a new combined model for cryptography and steganography. The model using the Hopfield Chaotic Neural Network (HCNN) for the cryptography which uses the chaotic trajectories of two neurons to produce main binary sequences for encryption the plain-text. The model used also the Double Density Discrete Wavelet Transform (DD DWT) to embed the cipher-text into the audio cover. Experimental results show that the model is efficiency and secure against the most knows attacks [16]. C. Multi-layer Neural networks Data security based on neural networks, 2005 Khaled M. presented a cryptography model based on the general regression neural network (GRNN). The proposed GRNN has three layers; each layer consists of a number of neurons, depending on the process type. The encryption 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 cryptology, 2010 Ilker D. proposed a chaotic cryptosystems based on artificial neural network and chaotic generator synchronization. The ANN model generated the chaotic dynamics by the numerical solution of Chua’s circuit. The proposed model have three initial conditions and time variable as input with two hidden layers and three chaotic dynamics as output. Many simulations are done on the number of neurons on hidden layer to find the best ANN structure obtained chaotic dynamics. The chaotic dynamics act as the key for encryption and decryption process. The ANN model does not have any synchronization problem. The difference between the chaotic dynamics can be considered as an advantage of the ANN based chaotic generator. The major weaknesses of analogue circuit and the numerical solution of chaotic circuit are eliminated with the proposed model. Simulation results show that the model are efficiency, secure, and can be applying on real time application [20].

The generation process of the key using back propagation neural network consists of three phases. The first phase is the feed-forward, the second phase is a backpropagation of the associated error and the third phase is related with weights adjustments. Simulation results show that the cryptosystem based on ANNs is very efficiency, and can be applied on hardware application [22]. Cryptography based on neural network, 2012 Eva Volna proposed multi-layer neural networks in cryptography. The multilayer neural networks modified by back-propagation. The planned model converted the input message intro ASCII code then gets the sequence of bit for each code which divided into 6 bit blocks are used as input for the encryption process. The cipher key is the neural network structure contained input layer, hidden layer, output layer, and updated weights. Experimental results show that the system is secure [23]. IV. SUMMARY In this paper, many of the important ANN techniques have been presented and analyzed. These techniques are based on:

AES Cryptosystem Development Using Neural Networks, 2011 Siddeeq. Y presented a new modification of the Advanced Encryption Standard (AES) to be immune against some attacks using nonlinear neural network. The neural network model performs cryptography processes via a symmetric key cipher that used as the initial weights for the neural network which trained to its final weight fast and low cost algorithm. The objective from the network has been selected to equivalent the output of the AES that have an efficient and recommended security. Simulation results show the proximity of the results accomplished by the proposed NN-based AES cryptosystem with that of the normal AES [21].

 The Tree Parity Machines (TPM) was used to 





Implementation of neural - cryptographic system using FPGA, 2011 Karam M. presented a stream cipher system based on pseudo Random Number Generator (PRNG) through using artificial Neural Networks (ANN). The PRNG model has a high statistical randomness properties for key sequence using ANN. The proposed neural pseudo random number generator consists of two stages; the first stage is generating a long sequence of patterns from perfect equation and initial value. So these patterns possess the randomness and unpredictable properties. The total number of equations and initial values depend on the number of bits that represented the initial value, the second stage is an artificial neural network (ANN) that gets the outputs of the previous stage and set it as input to the NN.



  

461

generate a secret key over the public channel on the output at each partner. The neural network with chaotic logistic map was used for cryptography by which both partners use the neural network as input for the logistic map, that generate the output bits to be learned. The General Regression Neural Network (GRNN) was used for encryption and decryption process based on three layers, where the input data divided into 3 bits and 8 bits as output. The training back-propagation neural network was act as public key, while Boolean algebra act as private key. The chaotic hopfield neural network with time varying delay was used to generate binary sequence for making plaintext, which considered as a random switching function for chaotic map. The neural network based on chaotic generator was used for generate chaotic dynamic act as a shard key. The initial weight value of the neural network was used after training as symmetric key. The Pseudo Random Number Generator (PRN) based on neural network was used in stream cipher as a key sequence generator.

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 chaotic sequence act as a triple key (combined of initial condition and control parameters) for cryptography.  The Layer Recurrent Neural Network (LRNN) was used to generate pseudo random number based on weight matrix obtained from layer weight of the LRNN.

[9]

[10]

[11]

V. CONCLUSION

[12]

In this paper, we summarize some recent researches about the application of neural network in the field of cryptography. The designed NN-based cryptosystem is a good idea of building very complicated cryptosystem, where the crypto analyst or the cracker not just need the topology of the NN and the key to crack the system, but also need to know the number of adaptive iterations and the final weights for the encryption and decryption systems. Applying higher numbers of plain-text/ cipher-text to the NN-based cryptosystem so as to make the error rate as minimum as possible. Lastly, according to the findings of this work, the trend for the years to come regarding the use of ANNs for cryptography tasks will be focused mainly on TPM, CNNs, and LRNN.

[13]

[14]

[15]

[16]

REFERENCES [1] [2]

[3] [4]

[5]

[6]

[7]

[8]

[17]

William Stallings, ―Cryptography and Network Security: Principles and Practice‖, (5th Edition), Prentice Hall, 2010. T. P.Wasnik , Vishal S. Patil , Sushant A. Patinge , Sachin R. Dave , Gaurav J. Sayasikamal, ―Cryptography as an instrument to network security‖, International Journal of Application or Innovation in Engineering & Management (IJAIEM), Vol. 2, Issue 3, 72-80, 2013. Wolfgang Kinzel, IdoKanter, ―Neural Cryptography‖, Proceedings TH2002 Supplement, Vol. 4, 147 – 153, 2003. Einat Klein, Rachel Mislovaty, IdoKanter, Andreas Ruttor, Wolfgang Kinzel, ―Synchronization of neural networks by mutual learning and its application to cryptography‖, In proceeding of: Advances in Neural Information Processing Systems 17 ,Neural Information Processing Systems NIPS, 2004. N. Prabakaran, P. Vivekanandan, ―A New Security on Neural Cryptography with Queries‖, Int. J. of Advanced Networking and Applications, Vol. 2, Issue. 1, 437-444, 2010. R. M. Jogdand, Sahana S. Bisalapur, ―Design of an efficient neural key generation‖, International Journal of Artificial Intelligence & Applications (IJAIA), Vol.2, No.1, 60- 69, 2011. Pratap Singh, Harvir Singh, ―Cryptography in structure adaptable digital neural networks", National monthly refereed journal of research in science & technology, Vol.1, Issue.12, 35-44, 2012. 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.

[18]

[19]

[20]

[21]

[22]

[23]

462

Wenwu Yu, Jinde Cao, ―Cryptography based on delayed chaotic neural networks‖, Physics Letters A, Vol. 356, (4) Elsevier, 333– 338, 2006. Jiyun Yang, Xiaofeng Liao, Wenwu Yu, Kwok-wo Wong, Jun Wei, ―Cryptanalysis of a cryptographic scheme based on delayed chaotic neural networks‖, Chaos, Solitons& Fractals, Vol. 40, Issue.2, 821825, 2009. Rajender Singh, Rahul Misra, Abhishek Chaudhary, ―Power consumption using artificial neural network in the field of cryptography‖, Journal of information, Knowledge and research in computer Engineering, Vol.2, Issue.2, 443- 446, 2012. Shweta B. Suryawanshi, Devesh D. Nawgaje, ―A triple-key Chaotic neural network for cryptography in image processing‖, International Journal of Engineering Sciences & Emerging Technologies, Vol. 2, Issue. 1, 46-50, 2012. Nitin Shukla, Abhinav Tiwari, ―An Empirical Investigation of Using ANN Based N-State Sequential Machine and Chaotic Neural Network in the Field of Cryptography‖, Global Journal of Computer Science and Technology Neural & Artificial Intelligence, Vol. 12, Issue.10,No. 1,17-26, 2012. Tariq A. fadil, Shahrul N. yaakob, Badlishah ahmad, abid yahya, ―Encryption of mpeg-2 video signal based on chaotic neural network‖, Journal of Engineering and Technology, Vol. 3, 35-42, 2012. Navita Agarwal, Prachi Agarwal, ―Use of Artificial Neural Network in the Field of Security‖, MIT International Journal of Computer Science & Information Technology, Vol. 3, No. 1, 42–44, 2013. B. Geetha vani, E. V. Prasad, ―A Hybrid Model for Secure Data Transfer in Audio Signals using HCNN and DD DWT‖, (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 4, No.7, 202-208, 2013. Khaled M. G. Noaman, Hamid Abdullah Jalab, ―Data security based on neural networks‖, Task Quarterly 9, No. 4, 409–414, 2005. Khalil Shihab, ―A Back propagation Neural Network for Computer Network Security‖, Journal of Computer Science 2 (9): 710-715, 2006. Seref S., NeclaÖ. , ―Neural Solutions for Information Security‖, Politeknik Dergisi, Journal of Polytechnic, Vol. 10, No. 1, 21-25, 2007. Ilker Dalkiran, Kenan Danis, ―Artificial neural network based chaotic generator for cryptology‖, Turk J Elec Eng& Comp Sci, Vol.18, No.2, 255- 240, 2010. Siddeeq. Y. Ameen, Ali H. Mahdi, ―AES Cryptosystem Development Using Neural Networks‖, International Journal of Computer and Electrical Engineering, Vol. 3, No. 2, 315- 318, 2011. Karam M. Z. Othman, Mohammed H. AL Jammas, ―Implementation of neural - cryptographic system using FPGA‖, Journal of Engineering Science and Technology, Vol. 6, No. 4, 411 – 428, 2011. Eva Volna, Martin Kotyrba, Vaclav Kocian, Michal Janosek, ―Cryptography based on neural network‖, Proceedings 26th European Conference on Modelling and Simulation, 2012.