SOFT COMPUTING BASED CRYPTOGRAPHIC TECHNIQUE USING KOHONEN'S SELFORGANIZING MAP SYNCHRONIZATION FOR WIRELESS COMMUNICATION (KSOMSCT)

International Journal in Foundations of Computer Science & Technology (IJFCST), Vol.4, No.5, September 2014
DOI:10.5121/ijfcst.2014.4508 85
SOFT COMPUTING BASED CRYPTOGRAPHIC
TECHNIQUE USING KOHONEN'S SELF-
ORGANIZING MAP SYNCHRONIZATION FOR
WIRELESS COMMUNICATION (KSOMSCT)
Arindam Sarkar1
and J. K. Mandal2
1
Department of Computer Science & Engineering, University of Kalyani, W.B, India
ABSTRACT
In this paper a novel soft computing based cryptographic technique based on synchronization of two
Kohonen's Self-Organizing Feature Map (KSOFM) between sender and receiver has been proposed. In this
proposed technique KSOFM based synchronization is performed for tuning both sender and receiver. After
the completion of the tuning identical session key get generates at the both sender and receiver end with the
help of synchronized KSOFM. This synchronized network can be used for transmitting message using any
light weight encryption/decryption algorithm with the help of identical session key of the synchronized
network. Exhaustive parametric tests are done and results are compared with some existing classical
techniques, which show comparable results for the proposed system.
KEYWORDS
Kohonen's Self-Organizing Map (KSOFM), soft computing, cryptography, Wireless Communication.
1. INTRODUCTION
A range of techniques are obtainable to protect data and information from attackers [1, 5, 6, 7, 8,
9, 10]. Existing TPM and PPM [2, 3, 4] technique have some limitations like secret seed values
used in the generation of identical input vector has to be transmitted to the other party via public
channel in the SYN frame in each iteration. This significantly increases the synchronization time.
Also for ensuring the security this parameters should not be transmitted via public channel.
Furthermore, TPM and PPM needs significant amount of synchronization steps. This may not
suitable in wireless communication because of resource constraints criteria. Proposed method of
this paper eliminates all the above stated drawbacks of the TPM and PPM. The proposed
technique performs the KSOFM based synchronization for generation of secret tuned session key
at both ends with fewer amounts of synchronization steps compared to TPM and PPM. The
organization of this paper is as follows. Proposed cryptographic technique has been discussed in
section 2. Experiments results of this technique are given in section 3. Conclusions are drawn in
section 4 and that of references at end.
2. THE TECHNIQUE
The proposed technique constructs the secret key at both sides using exchanged information. For
ensuring the randomness in every session, certain parameters value get randomly change in every
session like seed value for generating random inputs and weights, number of iteration to train the
map, different mathematical functions (Radial basis, Gaussian, Mexican Hat) for choosing the
random points from the KSOFM.

86
2.1 Synchronization Methodology
In this section, a novel Kohonen Self-Organizing Feature Map (KSOFM) based synchronization
of both sender and receiver machine has been presented. In this proposed technique input
examples are used to build the map through vector quantization method along with unsupervised
competitive learning for performing the synchronization. Proposed method uses unsupervised
learning method to represent input space of the training samples in a discrete 2D maps.
Neighborhood of each neuron (i.e. the connections of the neuron with adjacent neurons) in the
map depends on the dimension of the map. 2D regular spacing in a hexagonal or rectangular grid
uses to arrange the neurons. Following sub sections discussed about detailed methodology used in
synchronization procedure.
2.1.1 Initialization of weight vector
In this proposed technique KSOFM comprises of neurons along with a weight vector for
each neuron having a dimension same as the dimension of the input vectors. Consider the input
vector  T
n
x
x
x
X ,...,
, 2
1
 and weight vector  T
n
w
w
w
W ,...,
, 2
1
 .This scheme initially, assigns a
weight vector to each neuron (point) by arbitrarily choosing a neuron (point) of the input space.
The value of the weights vector is set to a tiny random numbers.
2.1.2 Selection of point
For initially assigning a weight vector to each neuron (point) an arbitrary neuron (point) ∈
get selected.
2.1.3 Unsupervised Training of KSOFM
The necessity of unsupervised learning mechanism in KSOFM is to produce similar response
from different parts of the network for a certain input patterns. This proposed technique uses
competitive learning in the training period to train the KSOFM. Euclidean distance between each
neuron and an arbitrary neuron (point) get calculated by:
     
 
2
2
2
2
2
1
1 ...
tan kn
n
k
k
k w
x
w
x
w
x
ce
Dis 






Where,
= 1,2, . . . ,
is the neuron number
kj is the entry of of the weight of neuron
The neuron (point) whose weight vector is most similar to the input is called the Best Matching
Unit (BMU). The weights of the BMU and neurons (point) close to it in the KSOFM lattice are
adjusted towards the input vector. The magnitude of the change decreases with time and with
distance (within the lattice) from the BMU. This proposed technique uses 2D KSOFM with 100
neurons. A learning rate  of 0.1 is used to train the KSOFM and decreasing the spread of the
neighborhood function by the rule:










iteration
of
no
Total
number
Iteration
_
_
_
_
1
0



87
Here, is initial spread. The value of decreases from the initial value to the final value (0)
constantly. Hence the neighborhood function influences all neurons of the map in the first time
and its influence on far neurons vanishes progressively. Near the end of the training only the
winner neuron will be updated so as to drive neurons to gravity centers.
2.1.4 Selection of Winner neuron
The winner neuron is a neuron which has a minimum distance to (arbitrary point). Winner
neuron gets selected based on the distance factor. The minimum distance winner
ce
Distan fulfills
the condition:
k
winner ce
Dis
ce
Dis tan
tan 
ℎ = 1,2, . . . ,
2.1.5 Updating of KSOFM
An updating rule has been applied over the entire map keeping in mind the priority of the
winner neuron and its closest neighbors. In this proposed technique updating is done using
following formula
  
old
k
old
k
new
k w
x
k
winner
Neighbor
w
w ,
,
, .
,
. 

 
Where  is the learning step and ℎ ( , ) is a neighborhood function with a
bell shape centered at the winner neuron. It is a function of the distance between the winner
neuron and the neuron k.
  2
2
, 
k
winner
e
k
winner
Neighbor



2.1.6 Reiteration on KSOFM
A new arbitrary point ∈ get selected and all the steps discussed in sub section
3.1.3 to 3.1.5 get perform again. This process is repeated for each input vector for a (usually
large) number of cycles .
2.2 Synchronization parameters
The spreading of the neighbourhood function ( ) is important since it controls the convergence of
the map. It should be large at the beginning and shrink progressively to reach a small value in
order to globally order the neurons over the whole map. The maximum value of
ℎ ( , ) = 1 corresponds to the winner neuron and value of Neighbor
function decreases when the distance between neurons and winner increases. Concerning the
value of the learning rate , it should be small enough to ensure the convergence of the KSOFM.
In this proposed synchronization technique the sender and receiver both uses the identical
KSOFM architecture along with identical parameters in each session. Following are the list of
parameters used in each session.
 Dimension of the KSOFM (2D or 3D)
 Number of neurons which specifies the number of different possible session keys
 Dimension of the weight vector specify the length of the key
 Seed value for generating random inputs and weights
 Number of iteration to train the map

88
 Different mathematical functions as a mask for choosing the random points from the
KSOFM (Radial basis, Mexican Hat, Gaussian etc.)
 Different index value for choosing different neurons (key) on the mathematical mask at
each session for forming the session key
Parameters that get negotiated at the initial stage of synchronization procedure between
sender and receiver by mutual agreement are completely random. By changing each of the
parameters randomly in each session security of proposed technique can be enhanced which
in turns decrease the success rate of the attackers.
In this proposed technique both A’s and B’s KSOFM are starts synchronization by exchanging
some control frames for negotiation of parameters value.
Synchronization (SYN) Frame:
2 4 1 2 16 2 4 4 16 (bits )
Figure 3: Synchronization (SYN) Frame
 Sender constructs a frame and transmitted to the receiver for handshaking purpose
in connection establishment phase. usually comprises of
: Following table illustrate the different command code in respect to
the different frame.
Table 1 Command Code
Command code Frame
00
01 _
10 _
11
_ : 4 bits _ is used to identify different SYN frame in different session.
: 1 bit is used to specify the dimension of KSOFM. Following table illustrate
the corresponds to the dimension.
DIM Index KSOFM Dimension
0 2D
1 3D
: 16 bits are used to illustrate the . Following table
shows the ℎ corresponds to the number of weights.
ℎ : 2 bits are used to illustrate the ℎ .
00
_
ℎ
(
ℎ )

89
Weight DIM Index Number of weights
00 64
01 128
10 192
11 256
: 2 bits are used to illustrate the . Following table illustrate
the different value corresponds to the different mathematical mask
function.
Mask Index Mathematical Mask Function
00 Mexican Hat
01 Gaussian
10 Radial Basis
11 Reserved
: 4 bits are used to illustrate the .
: 4 bits are used to illustrate the which is
the total number of neurons
( ℎ ): 16 bits are used in CRC.
 When the receiver receives the frame , the receiver should carry out integrity test.
 Receiver performs Integrity test after receiving the frame. If the messages are
received as sent (with no replication, incorporation, alteration, reordering, or replay) the
receiver will execute the synchronization phase.
Acknowledgement of Synchronization (ACK_SYN) Frame:
4 8 16 (bits)
Figure 4: Acknowledgement of Synchronization ( _ ) Frame
_ frame send by the receiver to the sender for positive acknowledgement of the
parameters value.
Negative Acknowledgement of Synchronization (NAK_SYN) Frame:
4 8 16 (bits)
Figure 5: Negative Acknowledgement of Synchronization ( _ ) Frame
0010
_
( ℎ )
0011
_
( ℎ )

90
_ frame send by the receiver to the sender for negative acknowledgement of the
parameters value.
Finish Synchronization (FIN_SYN) Frame:
4 8 16 (bits)
Figure 6: Finish Synchronization ( _ ) Frame
Finally, the receiver should send the frame _ to alert the sender.
2.3 Synchronization Algorithm
Input : Assign a weight vector to each neuron by arbitrarily choosing a point of the
input space
Output : Synchronized KSOFM
Method : The process operates on sender’s and receiver’s Kohonen's Self-
Organizing Feature Map (KSOFM) and generate synchronized session
key.
Step 1. Randomize the map's nodes' weight vectors
Step 2. Select an arbitrary input vector
Step 3. Traverse each node in the map
Step 3.1 Use the Euclidean distance formula to find the
similarity between the input vector and the map's
node's weight vector
Step 3.2 Track the node that produces the smallest distance
(this node is the best matching unit, BMU)
Step 4. Update the nodes in the neighborhood of the BMU (including the
BMU itself) by pulling them closer to the input vector
Step 5. Increase s and repeat from step 2
After a mutually predetermined iteration steps both sender and receiver stop their iteration. Now,
both sender and receiver have the identical final KSOFM because they have started with same
initial configuration and proceeds with same mutually agreement parameters. In this situation the
sender uses mathematical function as a mask to form the session key. The receiver would use the
same mutually pre agreed mathematical function as a mask in the map for reconstruct the session
key. Use of mathematical mask increases the security of the scheme because instead of one single
neuron, key can be constructed using several neurons on the mask. Also changing the mask
parameters several key can be generated. From this discussion it can be concluded that initially
mask method get slightly more amount of time but once the mask get set it takes less amount of
time. An adversary could not find the key because they do not have the map, mathematical
function for masking and other mutually pre agreed parameters. Mask method associate several
neurons around the designated neuron to form the key. This provides a significant improvement
to the security of the generation of session key. A mask hides all neurons other than the winner.
In this proposed scheme different mask functions like Gauss, Radial basis, Mexican hat functions
are get used randomly in different session. The winner neuron fixes the centre of the mask and
0001
_
( ℎ )

91
each neuron around the winner will be weighted and summed to the winner. The result is a
different key depending on the shape of the mask. The relation of the key is obtained by:
 
j
i
f
w
Key winner
j
i
q
ij
q
u ,
.
,
,
, 

= 1,2,… ,
Where,
 
j
i
f winner , is the mask centred at the winner neuron,
q
ij
w , , is the component of the vector associated to neuron ( , ) and
q
u
Key , , is the component of the ultimate (final) key.
A general form of the mask is represented by the the following equation:
 
2
2
2
2
1
2
.


w
k
w
k
winner be
e
a
x
f






with , , 1 , 2 , a neuron in the KSOFM, is the winner neuron. A huge number of
masks could be generated by changing parameters , , 1 , 2. Using the mask incontestably
enhances the security of the key. In wireless communication, instead of starting from the initial
state of the KSOFM key generation procedure a user may use the same trained KSOFM in
different session with different users by changing only the parameters value of the mask or mask
function used to determine the ultimate key. This procedure helps to save the resources of
wireless communication very efficiently.
3. SECURITY ANALYSIS
In this paper a KSOFM synchronized cryptographic technique has been proposed. The technique
generates the synchronized session key by tuning KSOFM of both sender and receiverThe
Following attacks are considered to ensure the robustness of the proposed technique.
Cipher text only Attack: In this type of attack, the attacker has access to a set of cipher text. In
cipher text only attack, encryption algorithm and cipher text is known to an attacker. An attacker
tries to break the algorithm or in simple words tries to deduce the decryption key or plaintext by
observing the cipher text. This proposed technique nullifies the success rate of this attack by
producing a robust KSOFM encrypted cipher text. The strength of resisting exhaustive key search
attack relies on a large key space. So, cipher text produces by this proposed methodology is
mathematically difficult to break. Thus a hacker has to try all such key streams to find an
appropriate one. Key stream have high degrees of correlation immunity. Thus it is practically
difficult to perform a brute-force search in a key-space.
Known Plaintext Attack: The attacker has access to one or more cipher text and some characters
in the original data. The objective is to find the secret key. Proposed technique offers better
floating frequency of characters. So, known plain text attack is very difficult in this proposed
technique.

92
Chosen Plaintext Attack: Here, the attacker has liberty to choose a plaintext of his/her choice and
get the corresponding cipher text. Since the attacker can choose plaintext of his/her choice, this
attack is more powerful. Again the objective of this attack is to find the secret key. This attack is
impractical because there is no obvious relationship between the individual bits of the sequence in
plain text and cipher text. So, it is not possible to choose a plaintext of his/her choice and get the
corresponding cipher text.
Chosen Cipher text Only Attack: The attacker can choose cipher text and get the corresponding
plaintext. By selecting some cipher text a cryptanalyst has access to corresponding decrypted
plaintext. Chosen cipher text only attack is more applicable to public key cryptosystems. This
technique has a good Chi-Square value this confirms good degree of non-homogeneity. So, it will
be difficult to regenerate plain text from the cipher text.
Brute Force Attack: A cryptanalyst tries all possible keys in finite key space one by one and
check the corresponding plaintext, if meaningful. The basic objective of a brute force attack is to
try all possible combinations of the secret key to recover the plaintext image and or the secret
key. On an average, half of all possible keys must be tried to achieve success but brute force
attack involves large computation and has a very high complexity. Due to high complexity brute
force attack will not be feasible. Proposed technique has a good entropy value near to equal 8
which indicates that brute force attack is very difficult in this proposed technique.
4. RESULTS
A total of 15 statistical tests of The NIST Test Suite have been performed to evaluate randomness
of the KSOFM synchronized session key proposed in this paper. These tests focus on a variety of
different types of non-randomness that could exist in a sequence. Some tests are decomposable
into a variety of subtests. The 15 tests are following:
Table 5 Statistical Results
Statistical Test
Expected
Proportion
Observed
Proportion
Status for
Proportion
of passing
P-value of
P-values
Status for
Uniform/
Non-uniform
distribution
Frequency
(Monobits) Test
0.972766 0.976329 Success 2.835246e-03 Non-uniform
Test for
Frequency
within a
Block
Runs Test 0.972766 0.975746 Success 0.321683e-01 Uniform
Longest Run of
Ones in a Block
0.972766 0.97051 Unsuccess 1.491737e+02 Uniform
Binary Matrix
Rank Test
0.972766 0.990000 Success 7.491904e-02 Uniform
Discrete
Fourier
Transform Test
0.972766 0.968329 Unsuccess 0.000000e+00 Non-uniform
Non-
overlapping
0.972766 0.988891 Success 0.000000e+00 Non-uniform

93
(Aperiodic)
Template
Matching Test
Overlapping
(Periodic)
Template
Matching Test
Maurer’s
“Universal
Statistical” Test
Linear
Complexity
Test
Serial Test 0.977814 0.973903 Unsuccess 0.000000e+00 Non-uniform
Approximate
Entropy Test
Cumulative
Sums Test
Random
Excursions Test
Random
Excursions
Variant Test
Table:6 Comparisons of 128 bit key length vs. average synchronization time (in cycle)
Key Length (128 bits) Average Synchronization time (in cycle)
KSOFM 2516,41
TPM (L=25) 2624,27
PPM 2811,04
Table shows KSOFM technique needs 2516,41 cycles in average to generate session key having a
length of 128 bit. Whereas existing TPM (L=25) and PPM needs 2624,27 and 2811,04 cycles
respectively, which larger than all the proposed techniques. KSOFM outperforms over existing
TPM and PPM. This is quite affordable in terms of resources available in wireless
communication.
Table:7 Comparisons of 128 bit key length vs. average synchronization time (in cycle) for group
synchronization (Group size=4)
Key Length (128 bits)
No. of Parties
participating in the
Group Synchronization
Average
Synchronization
time (in cycle)
KSOFM 4 15098,46
TPM (L=25) 4 15745,62
PPM 4 16866,24

94
Figure 16: Comparisons of 128 bit key length vs. average synchronization time (in cycle) for group
Table and figure shows KSOFM, TPM, PPM technique needs (15098,46), (15745,62),
(16866,24) cycles respectively in average to generate session key having a length of 128 bit for
synchronize group of 4 parties . This clearly indicates that proposed technique outperforms than
all other existing techniques at the time of group synchronization.
Table:8 Comparisons of 192 bit key length vs. average synchronization time (in cycle)
Key Length (192 bits) Average Synchronization time (in Cycle)
KSOFM 3173,41
TPM (L=25) 3347,15
PPM 3571,48
Table and figure shows KSOFM technique needs 3173,41 cycles in average to generate session
key having a length of 128 bit. Whereas existing TPM (L=25) and PPM needs 3347,15 and
3571,48 cycles respectively, which larger than the proposed techniques. KSOFM outperforms
over existing TPM and PPM. This is quite affordable in terms of resources available in wireless
communication.
Table: 9 Comparisons of 192 bit key length vs. average synchronization time (in cycle) for group
No. of Parties
Group Synchronization
Average
Synchronization
time (in Cycle)
KSOFM 4 19040,46
TPM (L=25) 4 20082,90
PPM 4 21428,88
14000
14500
15000
15500
16000
16500
17000
KSOFM TPM (L=25) PPM
Average Synchronization time (in Cycle)
Average Synchronization time
(in Cycle)

95
Table shows KSOFM technique needs 19040,46 cycles in average to generate session key having
a length of 192 bit for synchronize group of 4 parties This clearly indicates that proposed
technique outperforms than all other proposed and existing techniques at the time of group
synchronization.
Table: 10 Comparisons of 256 bit key length vs. average synchronization time (in cycle)
Key Length (256 bits) Average Synchronization time (in Cycle)
KSOFM 4719,72
TPM (L=25) 4851,86
PPM 5193,03
Table and figure shows KSOFM technique needs 4719,72 cycles in average to generate session
key having a length of 128 bit. Whereas existing TPM (L=25) and PPM needs 4851,86 and
5193,03 cycles respectively, which larger than the proposed techniques. This is quite affordable
in terms of resources available in wireless communication.
Table: 11 Comparisons of 256 bit key length vs. average synchronization time (in cycle) for group
No. of Parties
Group
Synchronization
Average
Synchronization
time (in Cycle)
KSOFM 4 28318,32
TPM (L=25) 4 29111,16
PPM 4 31158,18
17000
18000
19000
20000
21000
22000
Average Synchronization
time (in Cycle)

96
Table shows KSOFM technique needs 28318,32 cycles respectively in average to generate
session key having a length of 192 bit for synchronize group of 4 parties . This clearly indicates
that proposed technique outperforms than all other proposed and existing techniques at the time
of group synchronization.
Figure 19: Key length vs. average number of iterations
From figure it has been observed that if the length of the session key get increased then the
increased of average synchronization steps is linear.
Figure 20: Comparisons of relative time spent in GC to generate 128 bit session key
26000
27000
28000
29000
30000
31000
32000
Average Synchronization
time (in Cycle)
0
1000
2000
3000
4000
5000
128 bit Key Length 192 bit Key Length 256 bit Key Length
Average Number of Cycles
Average Number of Cycles
2.57
2.58
2.59
2.6
2.61
2.62
2.63
KSOFM TPM PPM
Relative Time Spent in GC (%)
Relative Time Spent in GC
(%)

97
From the above table and figure it has been shown that increasing order sequence of relative time
spent in GC in group synchronization phase is KSOFM, TPM and PPM.
Figure 21: Comparisons of number of Threads required generating 128 bit session key
From the above table and figure it has been shown that increasing order sequence of number of
thread required in group synchronization phase is KSOFM, TPM and PPM.
Figure 22: KSOFM dimension vs. average number of iterations
Figure shows the average number of iterations to be needed for generating 128, 192 and 256 bit
session key in 2D and 3D KSOFM. The above figure depicts that 3D KSOFM takes more
iterations to train the map in compared to 2D KSOFM. So, the energy consumption is more in 3D
KSOFM than 2D. For this reason 2D KSOFM is the best alternative in wireless communication
where resource constrains (in terms of energy, memory) is a vital issues for generation of session
key.
(a) (b)
(c) (d)
16
17
18
19
20
21
22
KSOFM TPM PPM
No. of Threads
No. of Threads
0
1000
2000
3000
4000
5000
6000
128 bit Key
Length
192 bit Key
Length
256 bit Key
Length
Average Number of Iterations
in 2D KSOFM
Average Number of Iterations
in 3D KSOFM

98
(e) (f)
(g) (h)
(i)
Figure 23: (a), (b), (c) shows the frequency distribution of plain text, TPM based encrypted text, proposed
chaos KSOFM based encrypted text, figure: (d), (e), (f) shows the floating frequency of plain text, TPM
based encrypted text, proposed chaos KSOFM based encrypted text, figure: (g), (h), (i) shows the
autocorrelation of plain text, TPM based encrypted text, proposed chaos KSOFM based encrypted text
5. CONCLUSION
Proposed technique is very simple and easy to implement in various high level language.
The test results also show that the performance and security provided by the proposed technique
is good and comparable to standard technique. The security provided by the proposed technique is
comparable with other techniques. To enhance the security of the technique, proposed technique
offers changes of some parameters randomly in each session. To generate the secret session key
index mask get exchanged between sender and receiver. This technique has a unique ability to
construct the secret key at both sides using this exchanged information. Since the encryption and
decryption times are much lower, so processing speed is very high. Proposed method takes
minimum amount of resources which is greatly handle the resource constraints criteria of wireless
communication. This method generates a large number of keys which is the same number of
neurons in the map. For ensuring the randomness in every session, some of the parameters get
change randomly at each session. proposed methods outperform than existing TPM, PPM and
does not suffers from Brute Force or Man-In-The-Middle (MITM) attack. No platform specific
optimizations were done in the actual implementation, thus performance should be similar over
varied implementation platform. The whole procedure is randomized, thus resulting in a unique
process for a unique session, which makes it harder for a cryptanalyst to find a base to start with.
This technique is applicable to ensure security in message transmission in any form and in any
size in wireless communication.

99
Some of the salient features of proposed technique can be summarized as follows:
a) Session key generation and exchange – Identical session key can be generate after the
tuning of KSOFM in both sender and receiver side. So, no need to transfer the whole
session key via vulnerable public channel.
b) Degree of security – Proposed technique does not suffers from cipher text only Attack,
known plaintext attack, chosen plaintext attack, Chosen cipher text only attack, brute
force attack.
c) Variable size key –128/192/256 bit session key with high key space can be used in
different session. Since the session key is used only once for each transmission, so there
is a minimum time stamp which expires automatically at the end of each transmission of
information. Thus the cryptanalyst will not be able guess the session key for that
particular session.
d) Complexity – Proposed technique has the flexibility to adopt the complexity based on
infrastructure, resource and energy available for computing in a node or mesh through
wireless communication. So, the proposed technique is very much suitable in wireless
communication.
e) Key sensitivity – Proposed method generates an entirely different cipher stream with a
small change in the key and technique totally fails to decrypt the cipher stream with a
slightly different secret session key.
f) Trade-off between security and performance – The proposed technique may be ideal for
trade-off between security and performance of light weight devices having very low
processing capabilities or limited computing power in wireless communication.
In future, some other soft computing based approach can be used to generate the session key.
ACKNOWLEDGEMENT
The author expresses deep sense of gratitude to the DST, Govt. of India, for financial assistance
through INSPIRE Fellowship leading for a PhD work under which this work has been carried out.
REFERENCES
[1] Atul Kahate, Cryptography and Network Security, 2003, Tata McGraw-Hill publishing Company
Limited, Eighth reprint 2006.
[2] R. Mislovaty, Y. Perchenok, I. Kanter, and W. Kinzel. Secure key-exchange protocol with an
absence of injective functions. Phys. Rev. E, 66:066102, 2002.
[3] A. Ruttor, W. Kinzel, R. Naeh, and I. Kanter. Genetic attack on neural cryptography. Phys. Rev.
E, 73(3):036121, 2006.
[4] Wolfgang Kinzel and ldo Kanter, "Neural cryptography" proceedings of the 9th
international
conference on Neural Information processing(ICONIP 03).
[5] Charles Pfleeger, Shari Lawrence Pfleeger, Security in computing, Third Edition 2003, pp 48,
Prentice Hall of India Pvt Ltd, New Delhi.
[6] Biham, E. and Seberry, J.”Py (Roo): A Fast and Secure Stream Cipher”. EUROCRYPT'05 Rump
Session, at the Symmetric Key Encryption Workshop (SKEW 2005), 26-27 May 2005.
[7] Chung-Ping Wu, C.C. Jay Kuo, “Design of Integrated Multimedia Compression and Encryption
Systems”, IEEE Transactions on Multimedia, Volume 7, Issue 5, Oct. 2005 Page(s): 828 – 839.
[8] HongGeun Kim, JungKyu Han and Seongje Cho.”An efficient implementation of RC4 cipher for
encrypting multimedia files on mobile devices”. SAC '07 Proceedings of the ACM symposium on
Applied computing, 2007, pp 1171--1175, NewYork, USA.

100
[9] Mantin and A. Shamir, “Weaknesses in the key scheduling algorithm of RC4”, Lecture Notes in
Computer Science, Vol. 2259, Revised Papers from the 8th Annual International Workshop on
Selected Areas in Cryptography, pp: 1 - 24, 2007.
[10] Sarkar Arindam, Mandal J. K., “Artificial Neural Network Guided Secured Communication
Techniques: A Practical Approach”, Paperback: 128 pages, Publisher: LAP LAMBERT Academic
Publishing (June 4, 2012), Language: English, ISBN-10: 3659119911, ISBN-13: 978-
3659119910.
Authors
Arindam Sarkar
INSPIRE FELLOW (DST, Govt. of India), MCA (VISVA BHARATI, Santiniketan,
University First Class First Rank Holder), M.Tech (CSE, K.U, University First Class
First Rank Holder).
Jyotsna Kumar Mandal
M. Tech.(Computer Science, University of Calcutta), Ph.D.(Engg., Jadavpur University)
in the field of Data Compression and Error Correction Techniques, Professor in
Computer Science and Engineering, University of Kalyani, India. Life Member of
Computer Society of India since 1992 and life member of cryptology Research Society of
India. Dean Faculty of Engineering, Technology & Management, working in the field of
Network Security, Steganography, Remote Sensing & GIS Application, Image
Processing. 25 years of teaching and research experiences. Eight Scholars awarded Ph.D.
and 8 are pursuing.

SOFT COMPUTING BASED CRYPTOGRAPHIC TECHNIQUE USING KOHONEN'S SELFORGANIZING MAP SYNCHRONIZATION FOR WIRELESS COMMUNICATION (KSOMSCT)

More Related Content

Similar to SOFT COMPUTING BASED CRYPTOGRAPHIC TECHNIQUE USING KOHONEN'S SELFORGANIZING MAP SYNCHRONIZATION FOR WIRELESS COMMUNICATION (KSOMSCT) (20)

More from ijfcstjournal (20)

Recently uploaded (20)

SOFT COMPUTING BASED CRYPTOGRAPHIC TECHNIQUE USING KOHONEN'S SELFORGANIZING MAP SYNCHRONIZATION FOR WIRELESS COMMUNICATION (KSOMSCT)