Professional Documents
Culture Documents
Thesis Submitted to
Acharya Nagarjuna University
FOR THE AWARD OF THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN
BY
A.V.N.KRISHNA
HYDERABAD
A.V.N.KRISHNA
Dr. A.VINAYA BABU, B.E, M.E (ECE), M.TECH (CSE), Ph. D (CS).
Professor of Computer Science & Engineering
DIRECTOR, ADMISSIONS,
CERTIFICATE
Abstract …
I
List of Tables …
ii
List of Figures …
Vi
1 Introduction …
1 – 08
2 Literature Review …
09 – 34
…
.
…
.
4.1 Training of the Algorithms using different keys & Time stamps …
86
4.2 Role of Statistical tests on Values generated by the algorithms under 100
study
5 Key Distribution Mechanism 104-106
5.1 Model for generation of Hash value for given key. 104
5.2 Example 105
5.3 Identifying any Garbled Key During Transmission 105
5.4 Conclusion 105
6 Comparative Study of Different developed …
107–122
Models
6.1 Computational power …
107
6.2 Complexity by Construction & Its Strength …
110
6.3 Avalanche Effect …
112
6.4 Differential & Linear crypto Analysis …
115
6.5 Security Analysis …
115
7 Applications of the Proposed Models …
123-126
8.1 Summary/Dissertation …
127
9 Future Work …
136
References
137-140
…
Research Work …
List of Tables
Table 3.2.1.3 Identifying the variations in cipher text by slight variations in the
key (Model 1)…...………………………………………………………………………49
Table 3.2.2.3 Identifying the variations in cipher text by slight variations in the
key (Model 2)………………………………………………………………….………...60
Table 3.2.2.4 Identifying the variations in cipher text by slight variations in the
plaintext(Model2)………….……………………………………………………………61
Table 3.2.3.3 Identifying the variations in cipher text by slight variations in the
Key (Model 3)………..………………………………………………………………….74
Table 3.2.3.4 Identifying the variations in cipher text by slight variations in the
plain text (Model 3)……………………………………………………………………..74
Table 3.2.3.5 Identifying multiple copies of cipher text for one plain text…………75
Table No. 3.3.1 Encryption & Decryption Process by Mathematical Model ..……81
Table 4. 1: Relationship between Random Key considered with the Basins (Sub
Keys) generated………………………………………………………………………..93
Table 4. 2: Level of Metric between Key and Basins (Sub Keys) generated for Sign
Function applied on the product of Modified ternary vector & the random matrix
key …………………………………………………………………………………......94
Table 4.3: Level of Metric between Key and Basins (Sub Keys) generated for
different functions applied on the product of Modified ternary vector & random key
.............………………………………………………………………………………….95
Table 4.4: Possible Relationship between Plain text and Cipher Text Generated in
Probabilistic encryption Algorithm (Model 4) (Depending on the random Key
Chosen)………………………………………………………………………………....96
Table 4.5: Conversion of plain text to cipher text for a 5 character Text……..…..97
Table 4.7: Conversion of plain text to cipher text for a 15 character Text…………97
Table 4.8: Conversion of plain text to cipher text for a 20 character Text…………98
Table 4.9: Multiple Cipher Texts for a 5 Character Plain Text …………………....98
Table 6.5.1: Comparative study of the proposed encryption algorithms like Model 1
& Model 2 with standard algorithm like DES in terms of computational overhead,
Data overhead, complexity and security analysis…………………………………...119
Table 6.5.2: Comparative study of the proposed encryption algorithms like Model 3
& Model 4 with standard algorithm like RC4 in terms of computational overhead,
Data overhead, complexity and security analysis…………………………………...120
List of Figures
INTRODUCTION
T
he necessicity of information security with in an organization have undergone
major changes in the past and present times. In the earlier times physical means
is used to provide security to data. With the advent of computers in every field,
the need for software tools for protecting files and other information stored on the
computer became important. The important tool designed to protect data and thwart
illegal users is computer security.
With the introduction and revolution in communications, one more change that
affected security is the introduction of distributed systems which requires carrying of data
between terminal user and a set of computers. Network security measures are needed to
protect data during their transmission. The mechanisms used to meet the requirements
like authentication and confidentiality are observed to be quite complex.
Security mechanisms usually involve more than a particular algorithm or protocol
for encryption & decryption purpose and as well as for generation of sub keys to be
mapped to plain text to generate cipher text. It means that participants be in possession of
some secret information (Key), which can be used for protecting data from unauthorized
use
rs. Thus a model has to be developed within which security services and
mechanisms can be viewed.
To identify and support the security services of an organization at its effective
level, the manager needs a systematic way. One approach is to consider three aspects of
information security that is Security attack, Security mechanism and Security services.
Security attack identifies different modes by which intruder tries to get unauthorized
information and the services are intended to counter security attacks, and they make use
of one or more security mechanisms to provide the service.
As the importance of information systems is ever growing in all most all fields,
electronic information takes on many of the roles, earlier they being done on papers. Few
information integrity functions that the security mechanism has to support are security
and confidentiality of the data to be transmitted and authentication of users.
There is no single mechanism that will provide all the services specified. But we
can identify a very important mechanism that supports all forms of information integrity
is cryptographic technique. Encryption of information is the most common means of
providing security. A model for encryption can be represented by the following Figure
1.1.
A crypto system is an algorithm, plus all possible plain texts, cipher texts and
keys. There are two general types of key based algorithms: symmetric and public key.
With most symmetric algorithms, the same key is used for both encryption and
decryption, as shown in Figure 1.2
The process of symmetric-key encryption can be very fast as the users do not
experience any significant time delay because of the encryption and decryption.
Symmetric-key encryption provides security to data as the key is shared only by the
participating parties. It also provides a degree of authentication, since information
encrypted with one symmetric key cannot be decrypted with any other symmetric key.
Thus, as long as the symmetric key is kept secret by the two parties using it to encrypt
communications, each party can be confident that it is communicating with the other as
long as the decrypted messages specify a meaningful sense.
Symmetric-key encryption will be successful only if the symmetric key is kept
secured by the two parties involved. If anyone else discovers the key, it affects both
confidentiality and authentication. The success of a symmetric algorithm rests in the key,
divulging the key means that any one could encrypt and decrypt messages. As long as the
communication needs to remain secure, the key must be protected between the
participating parties.
Public Key algorithms use two keys, one key for encryption and the other for
decryption. One key can be called as public key which can be declared public and the
other one is private that is, the key is known only to the particular participating party.
And also public key cryptography can be used for digital signing as it supports
authentication of users. The information encrypted with one key will only be decrypted
with the other key. Further more the decryption key cannot be calculated from the
encryption key. Figure 1.3 shows a simplified view of the way public-key encryption
works.
An attempted crypto analysis is called an attack. There are five types of attack. Each of
them assumes that the crypto analyst has complete knowledge of the encryption
algorithm used.
1. Cipher text only attack: Here the intruder is in hold of cipher text only. The
crypto analyst has cipher text of several messages, all of which have been encrypted
using the same encryption algorithm. The crypto analyst’s job is to recover the plain text
or the key used to encrypt the messages, in order to decrypt other part of messages
encrypted with the same keys.
2. Known Plaintext attack: The crypto analyst is in possession of pairs of known
plain text and cipher text. His job is to get the key used to encrypt the messages or an
algorithm to decrypt any messages encrypted with the same key.
3. Chosen Plaintext Attack (CPA): Here the crypto analyst is in hold of not only
cipher text but also parts of chosen plain text. Here the intruder is identified to be placed
at encryption site to do the attack. Differential crypto analysis is an example of this mode.
4. Chosen cipher text attack (CCA): Under the CCA model, the crypto analyst
is in possession of chosen cipher text and corresponding plain text being decrypted from
the private key. After it has chosen the messages, however, it only has access to an
encryption machine.
5. Chosen text: In this model, the analyst posses the encipher algorithm, Cipher
text to be decrypted, chosen plain text messages and corresponding cipher texts,
fabricated cipher text with the corresponding decrypted plain texts developed by the
private key.
Present work:
In this work an attempt has been made to generate two algorithms which provide
security to data transmitted. The first algorithm considers a random matrix key which on
execution by a series of steps generates a sequence. This sequence is used a sub key to
build three different encryption models. Each model can be used for encryption of data.
The second algorithm considers not only the key but also initialization vector and a time
stamp to generate sub keys which are used for encryption process. And also a mechanism
has been discussed which identifies any garbled key while transmitted from the Key
Distribution Centre.
In this work both the algorithms are discussed in terms of computational security,
computational complexity and computational overhead. Both the algorithms are studied
for their strengths and limitations. A crypto analytical study of the algorithms with
emphasis on probabilistic encryption is also considered in this study.
The encryption algorithms are compared with standard algorithms like RC4 and
DES. The algorithms are also discussed in terms of its applications and also about their
advantages and limitations in network security environment.
Chapter 2
LITERATURE SURVEY
A crypto system [1,7,8,17] is an algorithm which include all possible plain texts,
cipher texts and keys. There are two general types of key based algorithms:
symmetric and public key.
Initially the key is passed through a permutation function. Then for each of the 16
rounds, a sub key is generated by a combination of left circular shift and permutation.
At each round of operation, the plain text is divided to two 32 bit halves, and the
fallowing operations are executed on 32 bit right halve of plain text. First it is expanded
to 48 bits using a expansion table, then X-ORed with key, then processed in substitution
tables to generate 32 bit output. This output is permuted using predefined table and
XORed with left 32 bit plain text to form right 32 bit pre cipher text of first round. The
right 32 bit plain text will form left 32 bit pre cipher text of first round.
Decryption uses the same algorithm as encryption, expect that the application of
sub keys is reversed. A desirable property of any encryption algorithm is that a small
change in either plain text or the key should produce a significant change in the cipher
text. This effect is known as Avalanche effect which is very strong in DES algorithm.
Since DES is a 56 bit key encryption algorithm, if we proceed by brute force attack, the
number of keys that are required to break the algorithm is 2 56 . But by differential crypto
47
analysis, it has been proved that the key can be broken in 2 combinations of known
41
plain texts. By linear crypto analysis it has been proved that, it could be broken by 2
combinations of plain text.
The DES algorithm is a basic building block for providing data security. To apply
DES in a variety of applications, four modes of operations have been defined. These four
models are intended to cover all possible applications of encryption for which DES could
be used. They involve using a initialization vector being used along with key to provided
different cipher text blocks.
2.1.1.1.1 Electronic Code Book (ECB) mode: ECB mode divides the plaintext
into blocks m1, m2, ..., mn, and computes the cipher text ci = Ei(mi). This mode is
vulnerable to many attacks and is not recommended for use in any protocols. Chief
among its defects is its vulnerability to splicing attacks, in which encrypted blocks from
one message are replaced with encrypted blocks from another.
2.1.1.1.2 Cipher Block Chaining (CBC) mode: CBC mode remedies some of
the problems of ECB mode by using an initialization vector and chaining the input of one
encryption into the next. CBC mode starts with an initialization vector iv and XORs a
value with the plaintext that is the input to each encryption. So, c1 = Ek(iv XOR m1) and
ci = Ek(ci-1 XOR mi). If a unique iv is used, then no splicing attacks can be performed,
since each block depends on all previous blocks along with the initialization vector. The
iv is a good example of a nonce that needs to satisfy Uniqueness but not
Unpredictability.
2.1.1.1.3 Cipher Feed-Back (CFB) mode: CFB mode moves the XOR of CBC
mode to the output of the encryption. In other words, the cipher text c1 = p1 XOR
Sj(E(IV)). This mode then suffers from failures of Non-Malleability, at least locally to
every block, but changes to ciphertext do not propagate very far, since each block of
ciphertext is used independently to XOR against a given block to get the plaintext.
These failures can be seen in the following example, in which a message m = m1
m2 ... mn is divided into n blocks, and encrypted with an iv under CFB mode to c1 c2 ...
cn. Suppose an adversary substitutes c'2 for c2. Then, in decryption, m1 = Ek(iv) XOR
c1, which is correct, but m'2 = Ek(c1) XOR c'2, which means that m'2 = m2 XOR c2
XOR c'2, since m2 = Ek(c1) XOR c2. Thus, in m2, the adversary can flip any bits of its
choice. Then m'3 = Ek(c'2) XOR c3, which should lead to random looking message not
under the adversary's control, since the encryption of c'2 should look random. But m4 =
Ek(c3) XOR c4 and thereafter the decryption is correct.
2.1.1.1.4 Output Feed-Back (OFB) mode OFB mode modifies CFB mode to
feed back the output of the encryption function to the encryption function without XOR-
ing the cipher text.
A homophonic DES is a variant of DES that map search plaintext to one of many
cipher texts (for a given key). In homophonic DES a desired difference pattern with the
cipher text will be suggested with some key values including the correct one, oppositely
wrong pairs of cipher text. For a difference pattern which 56-bit plaintext to a 64-bit
cipher text using a 56-bit key. In this scheme, eight random bits are placed in specific
positions of the 64-bit input data block to maximize diffusion.
For example, the random bits in HDESS are the bit- positions 25, 27, 29, 31, 57,
59, 61 and 63. In this algorithm, after the initial permutation and expansion permutation
in the first round, these eight random bits will spread to bits 2, 6, 8, 12, 14, 18, 20, 24, 26,
30, 32, 36, 38,42,44,48 of the 48-bit input block to the S-boxes and will affect the output
of all the S-boxes. The 48 expanded bits must be exclusive-or’d with some key before
proceeding to the S-boxes, thus two input bits into the S-boxes derived from the same
random bit may have different values. This says that the random bits do not regularize the
input to the S-boxes, that is, the property of confusion does not reduce while we try to
maximize diffusion.
The decryption of the homophonic DES is similar to the decryption of DES. The
only difference is that eight random bits must be removed to get the original plaintext (56
bits). A homophonic DES can easily be transformed into a triple-encryption version by
concatenating a DES decryption and a DES encryption after the homophonic DES.
Security analysis: Thus there is a probability of 1/256 between a pair of texts. The
differential crypto analysis is also difficult on this mechanism. The diffusion of bits is
also more in this mode. Thus this mechanism provides some probabilistic features to DES
algorithm which makes it stronger from differential and linear crypto analysis.
2.1.1.4 AES:
The Advanced Encryption Standard (AES) was chosen in 2001. AES is also an
iterated block cipher, with 10, 12, or 14 rounds for key sizes 128, 192, and 256 bits,
respectively. AES provides high performance symmetric key encryption and decryption.
2.1.1.5.1 The Vernam Cipher: A Vernam cipher maps plaintext data with a
pseudo-random sequence to generate cipher text. Since each ciphertext element from a
Vernam combiner is the (mod 2) sum of two unknown values, the plaintext data is
supposed to be safe. But this mode is susceptive to several cryptanalytic attacks,
including known plain text and cipher text attacks. And if the confusion sequence can be
penetrated and reproduced, the cipher is broken. Similarly, if the same confusion
sequence is ever re-used, and the overlap identified, it becomes simple to break that
section of the cipher.
2.1.1.6 Nonces
A nonce [34] is a bit string that satisfies Uniqueness, which means that it has not
occurred before in a given run of a protocol. Nonces might also satisfy Unpredictability,
which effectively requires pseudo-randomness: no adversary can predict the next nonce
that will be chosen by any principal. There are several common sources of nonces like
counters, time slots and so on.
Ronald Rivest of RSA developed the RC4 algorithm, which is a shared key
stream cipher algorithm requiring a secure exchange of a shared key. The algorithm is
used identically for encryption and decryption as the data stream is simply XORed with
the generated key sequence. The algorithm is serial as it requires successive exchanges of
state entries based on the key sequence. Hence implementations can be very
computationally intensive. In the algorithm the key stream is completely independent of
the plaintext used. An 8 * 8 S-Box (S0 S255), where each of the entries is a permutation
of the numbers 0 to 255, and the permutation is a function of the variable length key.
There are two counters i, and j, both initialized to 0 used in the algorithm.
2.1.2.1.1 Algorithm Features: 1.It uses a variable length key from 1 to 256
bytes to initialize a 256-byte state table. The state table is used for subsequent generation
of pseudo-random bytes and then to generate a pseudo-random stream which is XORed
with the plaintext to give the cipher text. Each element in the state table is swapped at
least once.
2. The key is often limited to 40 bits, because of export restrictions but it is
sometimes used as a 128 bit key. It has the capability of using keys between 1 and 2048
bits. RC4 is used in many commercial software packages such as Lotus Notes and Oracle
Secure.
3. The algorithm works in two phases, key setup and ciphering. During a N-bit
key setup (N being your key length), the encryption key is used to generate an encrypting
variable using two arrays, state and key, and N-number of mixing operations. These
mixing operations consist of swapping bytes, modulo operations, and other formulas.
2.1.2.1.2 Algorithm Strengths: The difficulty of knowing which location in the
table is used to select each value in the sequence. A particular RC4 Algorithm key can be
used only once and Encryption is about 10 times faster than DES. Algorithm Weakness:
One in every 256 keys can be a weak key. These keys are identified by cryptanalysis that
is able to find circumstances under which one of more generated bytes are strongly
correlated with a few bytes of the key.
Thus some symmetric encryption algorithms have been discussed in this chapter.
They varies from block ciphers like DES, Triple DES, Homomorphic DES to stream
ciphers like RC4. To the symmetric encryption mechanisms concepts like application of
Nounce and dynamic substitution are discussed which provides randomness to the
encryption mechanism. This probabilistic nature to the encryption mechanism provides
sufficient strength to the algorithms against Chosen Cipher text attacks(CCA). The
security with all these mechanisms lies with proper sharing of keys among the different
participating parties.
The scheme shown in Figure 1.2 says public key is distributed and encryption
being done using this key. In general, to send encrypted data, one encrypt’s the data with
the receiver’s public key, and the person receiving the encrypted data decrypts it with his
private key.
As it happens, the reverse of the scheme shown in Figure 1.2 also works: data
encrypted with one’s private key can be decrypted only with his public key. This may not
be an interesting way to encrypt important data, however, because it means that anyone
with receiver’s public key, which is by definition published, could decipher the data. And
also the important requirement with data transfer is authentication of data which is
supported with Asymmetric encryption schemes, which is an important requirement for
electronic commerce and other commercial applications of cryptography.
2.2.1 Key Length and Encryption Strength:
Other ciphers, such as those used for symmetric key encryption, can use all
possible values for a key of a given length, rather than a subset of those values. Thus a
128-bit key for use with a symmetric-key encryption cipher would provide stronger
encryption than a 128-bit key for use with the RSA public-key encryption cipher.
This says that a symmetric encryption algorithm with a key length of 56 bits achieve a
equal security to Asymmetric encryption algorithm with a key length of 512 bits,
Encryption
Sender A does the following:-
Decryption
Recipient B does the following:-
1. Consider his own private key (n, d) to compute the plain text m = c^d mod n.
2. Convert the integer to plain text form.
Signature verification
Recipient B does the following:-
1. Using the sender’s public key, the received message digest is decrypted. From the
received message, the receiver independently computes the message digest of the
information that has been signed.
2. If both message digests are identical, the signature is valid.
Knapsack-based cryptosystems [2] had been viewed as the most attractive and the
most promising asymmetric cryptographic algorithms for a long time due to their NP-
completeness nature and high speed in encryption/decryption. Unfortunately, most of
them are broken for the low-density feature of the underlying knapsack problems. To
improve the performance of the model a new easy compact knapsack problem and
propose a novel knapsack-based probabilistic public-key cryptosystem in which the
cipher-text is non-linear with the plaintext.
These conditions imply that T1 is known in terms of T2. The equation for I=2, is a
relation between T1, T2 & T3. But since T1 can be expressed in terms of T2 , this
relation reduces to a relation between T2 and T3. This process of substitution can be
continued until Tn-1 can be formally expressed as Tn. But since Tn is known we can
obtain Tn-1.This enables us to begin back substitution process in which Tn-2,Tn-
3………….T3,T2 can be obtained.
For this tridiogonal system , it is easy to modify the Gaussian elimination procedures to
take advantage of zeros in the matrix of coefficients.
Referring to the tridiogonal matrix of coefficients above, the system is put into a upper
triangular form by computing new Ai.
Ai = Ai – (C i-1 /Ai)* Bi where i = 2,3……………ni. (eq. 6)
Di= Di – (C i-1 /Ai) * Di (eq. 7)
Then computing the unknowns from back substitution
Tn = Dn / An. (eq. 8)
Then Tn = Dk – Ak * T k+1 / Ak, k= ni-1, ni-2…3,2,1. (eq. 9)
Thus a Sequence of values are generated using tridiogonal matrix algorithm which can
be used as sub key in cryptographic techniques.
Introduction
The following sequence of steps identifies the methodology adopted in this work.
1. Definition of the problem.
2. Algorithms for generation of subkeys.
3. An algorithm 1 which multiplies ternary vector and a random matrix key
to generate a sequence. Dividing the sequence generated in algorithm 1
into basins based on equality of values. Mapping of the sequence or basins
to the plain text to generate cipher text. This mapping develops 3 models
which are discussed in detail as subunits. The developed algorithm is
trained to find an optimal key.
4. An algorithm 2 which considers a key, a time stamp and an Initialization
vector to generate sub keys which are mapped to plain text to generate
cipher text.
5. Training of the developed algorithms with different keys.
6. Adopting a suitable mechanism to identify any garbled keys while
transmission from the Key distribution centre.
7. Comparative study of the developed algorithms in terms of Computing
power, their Complexity in terms construction & strength, Avalanche
effect & Security analysis.
8. Comparative study of the algorithms with standard models like DES &
RC4.
9. Summary & Conclusion of the work.
3.1 Problem Definition
1. The type of operations used for transforming plain text to cipher text. All
encryption algorithms are based on two general principles. Substitution in which
each element in the plain text is mapped to another element and transposition in
which the elements in the plain text are re arranged. Most systems involve
multiple stages of substitution and transpositions.
2. The number of keys used. If the sender and receiver use the same key, the
system is referred as symmetric, single key, secret key or conventional
encryption. If the sender and the receiver each uses a different key, the system is
referred to as asymmetric, two key or public key encryption.
3. The way in which plain text is processed. A block cipher processes the input
one block of elements at a time, producing an output block for each input block.
A stream cipher processes the input element continuously, producing output one
element at a time, as it goes along.
Present work: In the present work two algorithms are developed. The first algorithm
uses a matrix key which on multiplication with a ternary vector and applying a sign
function on the product generates a sequence. This sequence will be used to generate
three different models of substitution technique. Thus the algorithm is considered to be a
substitution algorithm which uses a single key to be shared by both the sender and
receiver, and the cipher processes the input element continuously, producing output one
element at a time. The new encryption algorithm is based on the concept of Poly
alphabet cipher which is an improvement over mono alphabet. Three models are
developed from the given algorithm [22, 23]. The first two models like Model 1 & Model
2 can be classified under block ciphers and Model 3 is a stream ciphers. Each model is
having its own advantages and limitations.
The second algorithm considers not only key but also initialization vector and a time
stamp to generate sub keys which are used for encryption process.
3.2 Algorithm for generating the sequence
1. Consider the sequence for 0 to n values where n is a positive integer.
2. Convert each element of the sequence into ternary form of a given digit
number.
3. Represent the values of step 2 in a matrix form of (n+1) * (digit number).
4. Subtract 1 from each element of the matrix specified in step 3.
5. Consider a random matrix key of size (digit number*digit number).
6. Multiply the output of step 4 with the output of step 5.
7. Convert all positive values of matrix to 1, negative values to -1 and zero by 0.
8. Add 1 to each element of output of step 7.
9. Convert ternary values of step 8 into decimal form. A sequence is generated.
Example
Step1:
Consider the sequence for n= 0 to 26 values.
Step2:
Convert the sequence to ternary form of a 3 digit number
i.e. 0 ------- 000
1-------- 001
2-------- 002
.
.
.
26-------- 222
Step3:
0
1
0 0
0 0
2
0 0
0 1 0
1
0 1
2
0
0 1
0 2
1
2
0 2
0 2
0
1 0
1 0 1
2
1 0
1 1 0
= 1
R 1 1
2
0
1 1
1 2
1
1 2
1 2 2
0
2 0
2 0 1
2
2 0
0
1
2 1
2 1
2
0
2 1
2 2
1
2 2
2 2 2
Step 4:
Subtract 1 from each element of the above matrix and the resulting matrix R is
− 1 − 1 − 1
− 1 − 1 0
− 1 − 1 1
− 1 0 − 1
− 1 0
0
− 1 1
− 1
0
− 1
1
− 1 0
1
1
− 1
1
− 1 − 1
0
0 − 1 0
− 1 1
0
0 0 − 1
= 0
R 0 0
1
− 1
0 0
0 1
0
0 1
0 1 1
− 1 − 1
1
1 − 1 0
− 1 1
1
− 1
0
1 0
1 0
1
− 1
1 0
1 1
0
1 1
1 1 1
Step5:
Consider a random matrix
2 − 6
3 3
5
A= 1
4 −2 − 3
Step6:
R= R X A
− 1 − 7 1
− 7 − 4 − 2
− 13 − 1 − 5
4 − 6 − 1
− 2 − 31 − 4
− 8 − 7
9 − 3
0
− 5
3 − 2 − 6
− 3 − 9
1
1 − 4 5
− 5 − 1 2
− 11 − 1
2
6 − 3 3
R = 0 0
0
− 6 − 3
11 1
3
− 2
5 − 2
1
− 1 4 − 5
3 − 1 9
− 3 2 6
− 9 3
5
8 7
2 4
0
3
− 4 1
13 5
6
1
7 2
4
1 7 − 1
Step7:
Convert all positive values to 1, negative values to -1 and zero to 0 of the resulting
matrix of step 6.
− 1 − 1 1
− 1 − 1 − 1
− 1 − 1 − 1
1 − 1 − 1
− 1 − 1 − 1
− 1 − 1
− 1
0
− 1
1
− 1 − 1
− 1
1
− 1
1
− 1 1
1
− 1 − 1 1
− 1 − 1
1
1 − 1 1
= 0
R 0 0
− 1 − 1
1
1
− 1
1
− 1
1 1
− 1 1 − 1
− 1 1
1
− 1 1 1
− 1 1
1
1
1
1 0
1 1
− 1 1
1
1
1 1
1
1 1
1 1 − 1
Step8: Add 1 to each element of the matrix R
0 0 2
0 0 0
0 0 0
2 0 0
0 0 0
0 1 0
2 0 0
2 0 0
0 2 0
2 0 2
0 0 2
0 2 0
2 0 2
=
R 1 1 1
0 2 0
2 0 2
2 2 0
0 2 0
2 0 2
0 2 2
0 2 2
2 1 2
2 2 2
0 2 2
2 2 2
2 2 2
2 2 0
Step9:
2. r [27] = 2 0 0 18 0 3 18 18 6 20 2 6 20 13 6 20 24 6 20 8
8 23 26 8 26 26 24.
4.1. n[0] ie. ‘0’ is compared with r[27] values. There is a match at r[1],r[2] & r[4].
Neglect already visited elements. Thus b(0)=[0,1,2,4).
4.2 Step 4.1 is repeated with other elements of basin ie. 1, 2 & 4 values. For
elements 1& 4, there is no match of values in r[27]. For element 2, there is a
match at r[10]. Thus the basin b(0) is updated to (0, 1, 2, 4&10).
5. The procedure is repeated for the next element of the sequence of step 1 which is not
visited earlier. The other basins formed are
b(1)=(3,5,6,7,8,9,11,12,14,15,17,18,19,20,21,23)
b(2)=(13)
b(3)=(16,22,24,25,26)
Mapping the sequence or the basins on the plain text to generate three
models of cipher text.
3.2.1 Model 1:
A new substitution technique
The algorithm [22, 34] considers a matrix key and executes a sequence of steps
which generates a sequence. Each block of plain text is replaced by summation of
alphanumerical value of plain text and the sequence generated to form cipher text. Thus
the cipher text obtained becomes very difficult to break with out knowing the key.
3.2.2 Model 2:
A New Variable Length Key Block Cipher Technique for Network Security in Data
Transmission[15].
The algorithm considers a random matrix key which on execution of a sequence
of steps generates a sequence. Based on the equality of values this sequence is being
divided into basins. Each basin represents one block of data. Depending on starting input
plain text character, corresponding basin is considered as a key. Each block of plain text
is replaced by summation of alphanumerical value of plain text and the sequence
generated to form cipher text. The procedure is repeated for certain plain text depending
on chosen value. Thus the cipher text obtained becomes very difficult to be broken with
out knowing the key. In a variable length encryption scheme, the key used is of variable
length. The scheme requires one time execution of the key schedule to generate all sub
keys prior to encryption. This provides for slight resource burden on the key agility.
3.2.3 Model 3:
A New probabilistic Stream encryption scheme
In public key encryption there is always a possibility of some information being
leaked out. Because a crypto analyst can always encrypt random messages with a public
key, he can get some information. Not a whole of information is to be gained here, but
there are potential problems which allow a crypto analyst to encrypt random messages
with public key. Some information is leaked out every time to the crypto analyst, he
encrypts a message. In probabilistic multiple cipher texts are generated for one plain text,
a cryptanalyst can not generate any information by chosen plain text and chosen cipher
text attacks.
The algorithm [23] considers a matrix key and executes a sequence of steps which
generates a sequence. Based on the equality of values this sequence is being divided into
basins. The basins with minimum values will be neglected. Remaining basins represent
the characters of the alphabet. Each plain text character is replaced by a set of basins.
Each basin value is replaced by random value from that basin to generate cipher text.
Thus multiple cipher texts will be generated for one plain text and any one cipher text can
be used for transmission of data. Thus the cipher text obtained becomes computationally
infeasible to be broken with out knowing the key.
Thus the algorithm generates a subkey which will be used for encryption and
decryption processes used to develop three models. Model 1 and Model 2 are block
ciphers and Model 3 is stream cipher. Encryption & Decryption Processes are inverse to
one another which make the same algorithm to be used for encryption and decryption
process with a difference of addition in Encryption with subtraction in Decryption
Process.
A NEW SUBSTITUTION BLOCK CIPHER
Model 1
3.2.1 A new Substitution cipher for data security.
The algorithm that is going to be discussed in this work will generate a Sequence.
The algorithm
m considers a matrix key and executes a sequence of steps which generates the
sequence. Each block of plain text is replaced by summation of alphanumerical value of
the plain text and the sequence generated to form cipher text. Thus the cipher text
obtained becomes computationally infeasible to break with out knowing the key
It can be seen that to extract the original information from the coded text is highly
impossible for the third person who is not aware of encryption keys and the method of
coding.
Even if the algorithm is known it is very difficult to break the code and generate
key, given the strength of the algorithm .Thus given a short response time through
internet communication, the algorithm is supposed to be safe.
Decryption
Cipher C v n 02 R L
Text
Alpha 12 31 23 02 27 21
Numeric
equivalent
Add 36 12 31 23 38 27 21
Key 2 0 0 18 0 3
Subtract 10 31 23 20 27 18
Plain Text A v n K R I
3.2.1.5 Computing Power
Total number of computations considered in the given model for converting plain text to
cipher text.
Computation one: Converting n=0:26 to ternary vector. Let it be r.
Computation two: Calculating r-1 and storing it in r.
Computation three: Multiplying r with the key considered.
Computation four: Applying sign function on the product. Store it in r.
Computation five: Calculating r+1
Computation six: Converting output ternary vector to integer form. Let this be s,
the sequence generated.
Computation seven: Converting plain text to alphanumerical value.
Computation eight: Adding alphanumerical value of plain text to sequence
generated.
Computation nine: Applying mod function on the output.
Computation ten: Converting the output to characters of the alphabet to get
cipher text.
Thus the total number of computations in the first proposed model is 10.
Table No. 3.2.1.3: Identifying the variations in cipher text by slight variations in the
key.
(Model 1)
Case Considered Key Considered Plain text considered Cipher text formed
Case 1 [2 5 -6; 3 1 3; 4 - a b c d e cb cv e
2 -3]
Case 2 [3 5 -6; 4 1 3; 5 - a b c d e bbc v e
2 -3]
Case 3 [1 5 -6; 2 1 3; 2 - a b c d e i c f x e
2 -3]
` Thus we can see that, by changing the key slightly, there are a lot of variations in
the cipher text which provides maximum avalanche effect to the algorithm. This provides
for maximum strength and security to the algorithm. But since the model is a simple
substitution algorithm, the plain text variations to cipher text variations is negligible.
3.2.1.8 Security Analysis:
The model uses a sign function on the product of ternary vector and a matrix key
to generate the sequence. The sign function converts all positive values to 1, negative
values to -1, and zero with 0.This sequence is substituted for plain text to generate cipher
text. Thus it is impossible to generate the matrix key from the known plain text and
cipher texts. Thus this model is free from differential crypto analysis.
But this model uses a simple substitution technique to generate cipher text; it is
some what susceptible to linear crypto analysis. The Key can not be gained and not a
whole of information can be gained, but part of information may be gained in this model.
This algorithm is completely free from cipher text only, type of attack. By the other
attacks, the key may not be retrieved but a part of plain text may be retrieved.
3.2.1.9 Conclusion:
In this work a ternary system with a 3 digit number is used. So the sub key
generated is a 3 3 ie a 27 digit number. By considering a ternary vector with a four digit
number or five digit number, the length of the sub key can be increased by 34, 35 which
increases the length of sub key generated. Similarly by considering n –ary vector the
length of the subkey generated can still be increased. Thus by increasing the length of
subkey, security of cipher system can be increased still further.
A NEW VARIABLE LENGTH KEY BLOCK CIPHER
Model 2
3.2.2 A New Variable Length Key Block Cipher Technique for Network
Security in Data Transmission
The algorithm that is going to be discussed in this work is going to consider a
random matrix key which on execution of sequence of steps generates a sequence. Based
on the equality of values this sequence is being divided into basins. Each basin represents
one block of data. Depending on starting input plain text character, corresponding basin
is considered as a key. Each block of plain text is converted to alphanumerical values
which are mapped with the sub key to generate cipher text. The procedure is repeated for
certain plain text depending on chosen value. Thus the cipher text obtained becomes very
difficult to be broken with out knowing the key.
Key Words: Cryptography, Variable length key, Encryption Algorithm, Example, Add
function.
3.2.2.3 Advantages
1. It is almost impossible to extract the original information.
2. Even if the algorithm is known, it is difficult to extract the matrix key.
3. Versatile to users. Different users of internet can use different modified
versions of the new algorithm.
4. As per basin values, the same character is substituted by different alpha
numerical value which provides more security for the message.
3.2.2.4 Example:
The sequence generated from the product of ternary vector & Matrix key
considered (Ref. pp 44 )
2 0 0 18 0 3 18 18 6 20 10 ……………………………. 26 24.
The basins that can be formed using this sequence are
b(0)=(0,1,2,4,10)
b(1)=(3,5,6,7,8,9,11,12,14,15,17,18,19,20,23)
b(2)=(16,21,22,24,25,26)
1. Encryption
Plain text A V N K R I
Alpha 10 31 23 20 27 18
Numeric
Equivalent
Key(Basin 3 5 6 7 8 9
Considered)
Add 13 36 29 27 35 27
Mod 36 13 0 29 27 35 27
Cipher Text D 0 T R Z R
2. Decryption
Cipher Text D 0 T R Z R
Alpha 13 0 29 27 35 27
Numeric
Equivalent
Key(Basin 3 5 6 7 8 9
Considered)
Subtract 10 -5 35 20 27 18
Add 36, if 10 31 35 20 27 18
negative
Plain text A V n K R I
After some part of plain text, the procedure is repeated.
3.2.2.5 Computing Power
Total number of computations considered in the given model for converting
plain text to cipher text.
Computation one: Converting n=0:26 to ternary vector. Let it be r.
Computation two: Calculating r-1.
Computation three: Multiplying r with the key considered.
Computation four: Applying sign function on the product. Store it in r.
Computation five: Calculating r+1
Computation six: Converting output ternary vector to integer form. Let this be s,
the sequence generated.
Computation seven: Generating different basins by placing equality of values of
the sequence in one basin.
Computation eight: Converting plain text to alphanumerical value.
Computation nine: Considering the first character of plain text and applying a
mod function of the order of number of basins formed.
Computation ten: Depending on the output of the mod function, the
corresponding basin is used as a key.
Computation eleven: Adding the key to the alphanumerical value of the plain
text.
Computation twelve: Applying mod function on the output.
Computation thirteen: Converting the output to characters of the alphabet to get
cipher text.
Computation fourteen: Repeating the procedure for chosen part of plain text.
Thus the total number of computations in the proposed model is 14.
b(0)=(0,1,2,4,10)
b(1)=(3,5,6,7,8,9,11,12,14,15,17,18,19,20,21,23)
b(2)=(16,22,24,25,26)
Case 2: By increasing the values of key by 1 at each row
3 5 −6
A = key = 4 1 3
5 −2 −3
Sequence generated
r= 1 0 0 18 0 0 18 18 3 20 2 6 20 13 6 20 24 6 23 8 8 26 26 8 26 26 25.
Thus the basins formed which are used as sub keys are
b(0)= (0,1,2,4,5,10)
b(1)= (3,6,7,8,9,11,12,14,15,17,18,19,20,23)
b(2)= (16,21,22,24,25,26);
Case 3:
By decreasing the key values by 1 at each row
1 5 −6
A = key = 2 1 3
2 −2 −3
Note: Considering the basins as keys, by neglecting the basins which contain minimum
number of values, say 1.
Table No. 3.2.2.3 Identifying the variations in cipher text by slight variations in the
key (Model 2)
Table No.3.2.2.4 Identifying the variations in cipher text by slight variations in the
plain text (Model 2)
Case considered SubKey considered Plain text Cipher text
as per the first
character of plain
text
Case 1 36789 abcde dhjln
Case 1 16 22 24 25 26 bbcde hortu
Case 1 0 1 2 4 10 cbcde aceho
We will see a lot of variations in the cipher text generated for slight variations in the plain
text. Thus it provides a maximum avalanche effect to the algorithm which provides for
more strength and security.
3.2.2.9 Conclusion:
In the given work, a ternary vector with a 3 digit number is used. By using a n-ary
vector the length of the vector can be increased. By increasing the length of the vector,
the number of basins may not be increased. But the number of values in each basin will
be increased which provides more strength to the developed model. It is also observed
that by slight variations in the key values and plain text, the number of basins formed and
number of values of each basin are varying in nature, which provides a better avalanche
effect. This provides more security and strength to the algorithm
A NEW PROBABILISTIC STREAM CIPHER
Model 3
3.2.3 A New probabilistic encryption scheme:
With probabilistic encryption algorithms, a crypto analyst can no longer encrypt
random plain texts looking for correct cipher text. Since multiple cipher texts will be
developed for one plain text, even if he decrypts the message to plain text, he does not
know how far he had guessed the message correctly. To illustrate, assume a crypto
analyst has a certain cipher text ci. Even if he guesses message correctly, when he
encrypts message the result will be completely different cj. He cannot compare ci and cj
and so cannot know that he has guessed the message correctly. Under this scheme,
different cipher texts will be formed for one plain text. Also the cipher text will always be
larger than plain text.
Consider key
Multiply it
with ternary
vector
Apply a sign
function on
the output.
A sequence is
generated.
Similar
values of
sequence are
stored in
basins.
Each
character of
plain text are
represented
by set of
basins.
Values from
basins are
randomly
represented to
form cipher
text.
3.2.3.2 The new algorithm has the following features
1. A new probabilistic stream cipher is developed.
2. A matrix is used as a key.
3. A sequence is generated by the product of matrix key & the ternary vector.
4. This sequence is used to generate basins based on equal values.
5. Each plain text character is replaced by random values from basins to generate
cipher text.
6. Coding method.
3.2.3.4 Example:
For n= 0 : 26 ;
Sequence generated r (ref. pp 43)=2 0 0 18 0 3 18 18 6 20 2 6 20 13 6 20 24 6
20 8 8 23 26 8 26 26 24.
The basins that can be formed using this sequences are
b(0)=(0,1,2,4,10)
b(1)=(3,5,6,7,8,9,11,12,14,15,17,18,19,20,21,23)
b(2)=(16,22,24,25,26)
Now considering a base value of 3 i.e. representing a character with random values from
3 basins, the total characters that can be represented in this alphabet is 3 3 =27.
For example if K=11, it can be represented as 1 0 2 = b1 b0 b2.
Table No. 3.2.3.1 Encryption & Decryption Process by Model 3.
Encryption
Plain text A S K
Alpha numeric 01 19 11
equivalent
Equivalent basins b(0)b(0)b(1) b(2) b(0) b(1) b(1) b(0) b(2).
Cipher text Cj b b p z a n p d x
Decryption
Cipher text C i a b h x d i l b z
Alpha numeric 1 2 8 24 4 9 12 2 26
equivalent
Equivalent basins b(0)b(0)b(1) b(2) b(0) b(1) b(1) b(0) b(2).
Alpha numeric 01 19 11
equivalent
Plain text A S K
3.2.3.5 Computing Power
Total number of computations considered in the given model for converting plain text to
cipher text.
Computation one: Converting n=0:26 to ternary vector. Let it be r.
Computation two: Calculating r-1.
Computation three: Multiplying r with the key considered.
Computation four: Applying sign function on the product. Store it in r.
Computation five: Calculating r+1
Computation six: Converting output ternary vector to integer form. Let this be s,
the sequence generated.
Computation seven: Generating different basins by placing equal values of the
sequence in one basin.
Computation eight: Converting plain text to alphanumerical value.
Computation nine: Representing each character by a set of basins based on
chosen base value.
Computation ten: Representing the basins by their corresponding values in a
random manner.
Computation eleven: Converting the output to characters of the alphabet to get
cipher text.
Thus the total number of computations in the first proposed model is 11.
Computation overhead for a 27 character key
1st computation: 27 calculations, 2nd computation: 27 calculations
Key considered: 9 character key, 3rd computation: 27*9 calculations, 4th computation:
27 calculations, 5th computation: 27 calculations, 6th computation: 27 calculations, 7th
computation: 27*27 calculations +27*27 calculations, 8th computation: 27 calculations,
9th computation: 81 calculations, 10th computation: 81 calculations, 11th computation:
81calculations + Number of computations for random choice of values from basins (k).
Thus the total computational overhead by the third model is 2106+k calculations
3.2.3.6 Complexity of the Model
thus the basins that can be formed using this sequences are
b(0)=(0,1,2,4,10)
b(1)=(3,5,6,7,8,9,11,12,14,15,17,18,19,20,21,23)
b(2)=(13)
b(3)=(16,22,24,25,26)
Case 2: By increasing the values of key by 1 at each row
3 5 −6
A = key = 4 1 3
5 −2 −3
Sequence generated
r= 1 0 0 18 0 0 18 18 3 20 2 6 20 13 6 20 24 6 23 8 8 26 26 8 26 26 25.
Thus the basins formed are
b(0)= (0,1,2,4,5,10)
b(1)= (3,6,7,8,9,11,12,14,15,17,18,19,20,23)
b(2)=(13)
b(3)= (16,21,22,24,25,26);
Case 3:
By decreasing the key values by 1 at each row.
1 5 −6
A = key = 2 1 3
2 −2 −3
Sequence generated by the model.
r= 11,1,3,20,0,6,18,18,6,20,2,6,20,13,6,20,24,6,20,8,8,20,26,6,23,25,15.
Table No. 3.2.3.4 Identifying the variations in cipher text by slight variations in the
plain text
Considering Plain text Cipher text Data overhead
different cases
Case 1 a b c 0biabpacj 3 times
Case 1 b b c b0v0jpj0d 3 times
Case 1 c b c 0s0bdxacj 3 times
Table No. 3.2.3.5 Identifying multiple copies of cipher text for one plain text
Considering Plain text Cipher text Data overhead
different cases
Case 1 a b c 0ahbap0kj 3 times
Case 1 a b c 0jwdavjrd 3 times
Case 1 a b c Jbnbdyatd 3 times
Thus a lot of variation in the cipher text is observed by slight variations in the key
and the plain text. This provides for maximum avalanche effect on the algorithm which
provides maximum strength and security to the algorithm. And also it is observed that by
slight variations in the key, there is a substantial variation in the number of basins formed.
This will reduce the data overhead as less combination of basins will be sufficient to
support the characters of the alphabet.
The model uses a sign function on the product of ternary vector and a matrix key
to generate the sequence. The sign function converts all positive values to 1, negative
values to -1, and zero with 0.This sequence is substituted for plain text to generate cipher
text. Thus it is impossible to generate the matrix key from the known plain text and
cipher texts. Thus this model is free from differential crypto analysis and since basins
with different and variable values are mapped to the characters of plain text to form the
cipher text, it is also free from linear crypto analysis.
This algorithm is completely free from cipher text, known plain text, chosen plain
text, chosen cipher text attacks. This algorithm is also free from public key attacks.
3.2.3.9 Conclusion:
In this work a ternary system with a 3 digit number is used. So the sequence
generated is a 27 digit number. By using a n-ary vector, the length of the vector can be
further increased. But this does not guarantee in the increase in number of basins formed.
The number of values of generated sequence will increase in each basin. Thus the new
model provides a new probabilistic substitution mechanism where each character of plain
text is replaced by two or three characters of cipher text depending on the chosen key.
And also the model develops multiple cipher texts for one plain text. The algorithm
provides almost equal security at low computational overhead. And also the given
algorithm is free from differential and linear crypto analysis, which makes it suitable in
data encryption.
The limitation with this algorithm is more data has to be transmitted than the
actual data which demands for more band width requirements.
3.3 A New Mathematical Model with a key & a Time stamp for
encryption mechanism.
In this work a model is going to be used which develops data distributed over a
identified value which is used as nonce (IV). By properly considering an empirical
value, data is derived from the developed model. This empirical value is considered
as a key. The process is repeated for different timings which are used as time stamps
in the encryption mechanism [11,33]. Thus this model generates a distributed
sequence which is used as sub key. The model involves a identified value which is
used as nonce (IV), a considered empirical value which is used as key and different
timing as time stamps which are very important parameters in symmetric data
encryption schemes which supports not only security but also timeliness of
encryption mechanism and also acknowledgement between the participating parties.
3.3.1 Mathematical modeling of the problem
The approach to time series analysis was the establishment of a mathematical model
describing the observed system. Depending on the appropriation of the problem a
linear or nonlinear model will be developed. This model can be useful to generate
data at different times to map it with plain text to generate cipher text.
(α/∂.x) *(T I+1 n+1- T I n+1 ) – (α/ ∂x)* (TI n+1- T I-1 n+1 ) = (∂ x /∂ t)* ( T I n+1 –T I n ).
(α*∂ t /(∂.x) 2 )*(T I+1 n+1- T I n+1 ) – (α*∂ t /(∂.x) 2 )*(TI n+1- T I-1 n+1 ) = ( T I n+1 –T I n
).
(-2*α*∂ t /(∂.x) 2 )*(T I n+1 ) + (α*∂ t /(∂.x) 2 )*( T I-1 n+1 ) + (α*∂ t /(∂.x) 2 )*(T I+1 n+1)
= ( T I n+1 –T I n ).
n+1
where T I represents data value for the considered grid point for the current delt,
n+1 n+1
T I-1 &T I+1 represents data values for the preceding and succeeding grid
points for the current delt, T I n represents data item for past delt, α be a constant.
Considering α which is a key for the given model, the coefficients are obtained for
each state (grid point) in terms of A(I) refers to data value of the corresponding grid
point, C(I) and B(I) refers to data values of preceding and succeeding grid points for
the current delt, D(I) refers to data value of the considered grid point in the preceding
delt.
A(I)= 1 + 2 α delt/(delx)**2.
B(I)= -α delt/(delx)**2.
C(I)= - α delt/(delx)**2.
D(I)= T I n
These conditions say that T1 is known in terms of T2. The equation for I=2, is a
relation between T1, T2 & T3. But since T1 can be expressed in terms of T2 , this
relation reduces to a relation between T2 and T3. This process of substitution can be
continued until T M-1 can be formally expressed as TM. But since TM is known we can
obtain T M-1.This enables us to begin back substitution process in which T M-2,
T M-3…T3, T2 can be obtained. This process is continued until further iterations cease
to produce any significant change in the values of T’s. Finally the data distribution is
obtained for all grid points for different times by considering a suitable α which is
used as key.
3.3.4 Results
By considering a suitable key α =4, del t= 2, delx =2 for a total time stamp of 6 units,
Different data values obtained are
For del t=2, time =2;
33 6 7 4 33 8 11 13 32 22 29 20 26 0 18 10 17 11 1 1;
Thus by using the same key, by changing the time stamp values different sequences
can be generated which are used as sub keys. These sub keys can be mapped to plain
text to generate cipher text.
Table No. 3.3.1: Encryption & Decryption process by Mathematical Model
Encryption
Plain Text A S K S
Conversion to 10 28 20 28
alpha numeric
value
Sub key 33 6 7 4
Total 43 34 27 32
Mod 36 07 34 27 32
Cipher Text 07 Y R W
Decryption
Cipher Text 07 Y R W
Conversion to 07 34 27 32
alpha numeric
value
Add 36 if less 43 34 27 32
than 9
Sub key 33 6 7 4
Subtract 10 28 20 28
Plain Text A S K S
3.3.5 Number of Computations for converting plain text to cipher text:
3.3.7 Avalanche effect: It mainly deals with identifying major variations in cipher text
by small variations in key values. In the given model it is observed that a small change in
key values, Time stamps or the nonce is providing major changes in cipher values.
12 8 11 19 30 17 7 14 22 9 4 16 28 26 22 9 25 7 10 6 0 12 27 1 35 22 7 15 23 2 19 7 15 8
25 20 31 14 6 2 25 0.
3.3.9 Conclusion:
This encryption mechanism uses a Initialization Vector, Time Stamp & Key to
generate distributed sequences which are used as sub-keys. Since the time stamp is
variable in nature, the model provides sufficient security against crypto analysis. The
model is free from cipher text attack, known plain text & cipher text attacks. In the
given model past & present time stamps have been used to generate data. By properly
guessing future time stamps, the model can be made still stronger.
Chapter 4
TRAINING OF THE DEVELOPED MODELS
T raining an encryption model involves variable length key, a plain text and an
algorithm which applies the key on the plain text to generate cipher text. This
allows for the data to be transmitted over the network in some form which cannot be read
by any intruder.
In the first algorithm, the model will be trained for different keys and an analysis of
the models will be done. The keys that are used for training on the model are 3*3 keys
and 4*4 matrix keys. By having small variations in the keys, the model will be studied for
its increase in performance and working. The model will also be trained for its increase in
security by having slight variations in the key values. The model will also be trained for
changes in data overhead for small variations in the key values.
The model will be trained for different input vectors applied on the matrix keys to
generate different sequences which will be used to generate basins. The model is trained
for different input vectors of 27, 81 values with base 3 and 64 & 100 values with a base
4. For example, if the input vector of 27 values ranging from(0-26) is considered, with a
base value of 3 & the number being represented in 3 digit form, then the values will be
represented as (0 0 0 to 2 2 2).
For example,
(26) 3 in 3 digit format = 2 *3 2 +2*3 1 +2*3 0.= ternary equivalent is ( 2 2 2) 3
(80) 3 in 4 digit format = 3*3 3+3*3 2+3*3 1+3*3 0 = ternary equivalent is (3 3 3 3) 3
(100) 4 in 4 digit format = 1*4 3+2*4 2+1*4 1+0*4 0. = 4-ary equivalent is (1 2 1 4) 4
(255) 4 in 4 digit format = 3*4 3+3*4 2+3*4 1+3*4 0.= 4-ary equivalent is ( 3 3 3 3) 4
For example,
Case 1: Considering the key A of size (3*3)
2 5 −6
A = key 3 1 3
4 − 2 −3
Sequence generated from the proposed model
n=0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
26
r=2 0 0 18 0 3 18 18 6 20 2 6 20 13 6 20 24 6 20
r= 8 8 23 26 8 26 26 24.
1 5 −6 1
2 1 3 2
A = key =
3 − 2 −3 3
4 2 4 4
Sequence generated for n=0 to 80 values represented to the base 3.
r=0,33,60,0,0,6,0,9,20,54,57,60,0,0,40,0,19,20,54,54,60,54,54,74,9,20,20,33,60,60,0,6,26
,9,20,26,57,60,61,0,40,80,19,20,23,54,60,71,54,74,80,20,20,47,60,60,71,6,26,26,20,26,26
,60,61,80,40,80,80,20,23,26,60,71,80,74,80,80,20,47,80.
The basins formed by considering similar values,
b(0)=(0,3,4,6,12,13,15,30,39,5,31,57,10,36)
b(1)= (9,54,7,24,33,18,19,21,22,45,48,1,27,16,42)
b(2)= (20,60,8,17,25,26,34,43,51,52,69,78,2,11,28,29,37,46,54,55,63,72,32,35,
58,59,61,62,71,9,18,19,21,22,45,48,38,64,47,56,73,7,24,33,16,42,53,79,1,27)
b(3)= (23,74,44,70,49,75)
b(4)= (40,14,66)
b(5)= (80,41,50,65,67,68,76,77).
Case 5:
2 5 −6 1
3 1 3 2
A= key =
4 − 2 −3 3
5 2 4 4
2 [3 5 -6; 4 1 3; 4 3 64 3 Low
5 -2 -3]
3 [1 5 -6; 2 1 3; 6 2 36 2 High
2 -2 -3]
4 [1 5 -6 1;2 1 3 6 2 36 2 High
2;3 -2 3 3;4 2
4 4]
5 [2 5 -6 1;3 1 3 4 3 64 3 Low
2;4 -2 -3 3;5
2 4 4]
6 [3 5 -6 1;4 1 3 4 3 64 3 Low
2;5 -2 -3 3;6
2 4 4]
7 [1 5 -6 1;2 1 3 6 2 36 2 High
2;3 -2 -3 1;4
2 4 4]
8 [2 5 -6 1;3 1 3 5 3 125 3 Medium
2;4 -2 -3 1;5
2 4 4]
9 [3 5 -6 1;4 1 3 5 3 125 3 Medium
2;5 -2 -3 3;6
2 4 4]
Table 4.3: Level of Metric between Key and Basins (Sub Keys) generated for
different functions applied on the product of Modified ternary vector & random key
1 2 2]
6 [1 2 3;3 2 Mod, 5 6 High
2;
1 2 2]
7 [1 2 3;3 2 Mod, 7 6 High
2;
1 2 2]
8 [11 8 9;12 Mod, 3 8 High
13 10;11 9
9]
9 [11 8 9;12 Mod, 5 7 High
13 10;11 9
9]
It is observed that the by using different functions like Modular arithmetic or Sign
functions on the product of Modified ternary vector and matrix key, the relation to key
and sub keys almost remains the same ie varies between 3 -8 basins(sub keys).
Table 4.4 : Possible Relationship between Plain text and Cipher Text Generated in
Probabilistic encryption Algorithm (Model 3) (Depending on the random Key
Chosen).
Thus depending on the key considered and the number of sub keys(Basins) generated,
one plain text character can be replaced by different combinations of cipher texts. As per
the above tables it is observed that if one plain text character is replaced by 3 basins, then
the number of possible cipher text characters for one plain text character is 30 to 450
combinations. For different possible keys different combinations of cipher texts are
possible for one plain text character which varies from a lesser value to higher values.
Table 4.6: Conversion of plain text to cipher text for a 10 character text
Case 2:
Plain Text for a 10 Character text The require
Cipher text for model 1 Vhesmqxjzf
Cipher text for model 2 Wmkymrvjjf
Cipher text for model 3 Xapapy0izyoedkveypxudqoayabdix
Table 4.7: Conversion of plain text to cipher text for a 15 character text
Case 3:
Plain Text for a 15 Character text The requirements
Cipher text for model 1 Vhesmqxjzfukptu
Cipher text for model 2 Wmkymqfsuz
Cipher text for model 3 Ydp0ppmbby00dtpnpyzo0sddvdcjsychrdsyrgyp0vxai
Table 4.8: Conversion of plain text to cipher text for a 20 character text
Case 4:
Plain Text for a The reqiurementsof inf
20 Character
text
Cipher text for Vhesmqxjzfukptuuhioi
model 1
Cipher text for Wmkymrvjjfqfsuzpnjwh
model 2
Cipher text for V0pdvyacpybal0yspxzkdlday00dqycchdoyelpyappdolpdbydqaasrxdxa
model 3
Thus multiple cipher texts can be formed for one plain text in probabilistic
encryption scheme which makes the algorithm strong from chosen cipher text attacks.
Training of second Algorithm:
In the Second algorithm, the model will be trained for different keys, time stamps &
Nonce values and an analysis of the model will be done. In the given model it is observed
that a small change in key values, Time stamps or the nonce is providing major changes
in cipher values.
12 8 11 19 30 17 7 14 22 9 4 16 28 26 22 9 25 7 10 6 0 12 27 1 35 22 7 15 23 2 19 7 15 8
25 20 31 14 6 2 25 0.
4.2 Role of Statistical tests on Values generated by the algorithms under study.
In the given study, two algorithms are discussed. Both the algorithms execute a series of
steps and generate a sequence. This sequence is being used as sub key to be mapped to
plain text to generate cipher text. The strength of the encryption & Decryption process
depends on the strength of sequence generated. In this part of work some statistical tests
like Uniformity tests, Universal tests & Repetition tests are tried on the sequence
Uniform test tries to study the uniformity of a sequence generated. From the sequence
generated, the first two consecutive points are considered as coordinates of a graph. The
process is repeated for all the values of the sequence to generate the graph. For example,
if a,b,c,d……..n are the values of sequence generated, then the coordinates of the graph
are (a,b),(c,d)……….(n-1,n). If they are dependent they will form into lines or they will
form into a plane. In this study, different cases are tried for both the algorithms to study
Universal tests: It tries to see whether the data can be compressed. Since in the given
sequence, the repetition of values is minimum, we can say that the developed sequences
Repetition test: This test studies that each character of the sequence is not identical to the
earlier character. In the sequence generated by the first algorithm, the values are random
and unique in nature, the probability that the values get repeated is 1/27 . In the second
algorithm even though some values of the sequence are repeated, the distribution of
values is not uniform and random in nature which does not give any insight about he
pattern of the sequence. Thus we can say that the sequence generated by both the
algorithms is relatively free from Uniformity test, Universal test & Repetition tests.
The analysis is being done on sequences generated by both the algorithms which are
represented as case studies in the earlier part of the work. The analysis is being
30 30
25 25
20 20
Scatterd Scattered
15 values for case 15 values of case
1 2
10 10
5 5
0 0
0 10 20 30 0 10 20 30
30 90
80
25
70
20 60
Scattered Scattered
50
15 values of values of case
40
Case 3 4
10 30
20
5
10
0 0
0 10 20 30 0 50 100
90 90
80 80
70 70
60 60
Scattered Scattered
50 50
values of case values of case
40 40
5 6
30 30
20 20
10 10
0 0
0 50 100 0 50 100
Fig 4.2: Statistical Representation of Data of Algorithm 2:
30 35
25 30
25
20
sequential 20 Sequential
15 values of alg data of alg 2
2: Case 1 15 Case 2
10
10
5 5
0 0
0 20 40 0 10 20 30
18 40
16 35
14 30
12 25
10 Sequential Sequential
values of alg 20 values of alg 2
8
2: case 3 15 Case 4
6
4 10
2 5
0 0
0 10 20 30 0 20 40
35
30
25
20 Sequential
values of alg
15 2: Case 5
10
5
0
0 20 40
From the graphs we can say that the data distribution is non uniform and random in
nature in the sequence generated by the algorithms which provides sufficient strength
against crypto analysis.
Chapter 5
Key distribution mechanism forms an integral part of encryption decryption process. The
success of Symmetric encryption lies with proper and confidential sharing of private key
among the participating parties. Some times keys get garbled in transmission. Since a
garbled key can mean mega bytes of unacceptable cipher text, this sis a problem. All keys
should be transmitted with some kind of error detection and correction bits. In this part of
work, a new approach is tried to identify any garbled form of key is transmitted to
participating parties. A hash value is generated from the key to be transmitted. The
generated hash value which is appended to the key encrypted by either private keys of
the participating parties or the master key shared by the participating parties provide
security & confidentiality to key transmitted and also identifies any garbling of key
values during transmission.
In this part of work, a new approach is tried to identify whether any garbled form of key
is transmitted to participating parties.
In the methodology of the proposed work, two algorithms are presented, the first
algorithm deals with generating the product of ternary vector and a random matrix key
and applying a sign function on the product to generate a sequence (pp 42). In the second
algorithm a mathematical model is considered which uses a key, time stamp and nonce
values to generate sequence (pp 84).
5.1 Model for generation of Hash value for given key.
5.2 Example:
Considering the first algorithm,
1. Generated sequence from the algorithm 0 13 11 6 2 12 10 5 1
2. Binary form of the sequence= 00000000 00001101 00001001 00000110 00000010
00001100 00001010 00000101 00000001.
3. Considering 6 bits as one block of data.
4. Generated hash value= 101111 = 47.
5.4 Conclusion:
This mechanism uses the same algorithm for the generation of sub keys and also for the
generation of sequence which is used to generate hash value. If any errors takes place in the
key while transmission, there will be a mismatch between received hash value & generated
hash value and accordingly the participating parties ask for retransmission of key. Thus this
model identifies any garbled form of keys while transmission from Key Distribution
Centre.
Chapter 6
COMPARATIVE STUDY OF DIFFERENT
DEVELOPED MODELS
Model 2:
Computational overhead for a 9 character key
1st computation: 27 calculations, 2nd computation: 27 calculations, 3rd computation:
27*9 calculations, 4th computation: 27 calculations, 5th computation: 27 calculations,
6th computation: 27 calculations, 7th computation: 27*27 calculations +27*27
calculations, 8th computation: 27 calculations, 9th computation: 03 calculations
depending on chosen rule, 10 computation 03 calculations, 11 computation: 27
calculations, 12th computation: 27 calculations, 13th computation: 27 calculations, 14th
computation: 03 calculations depending on the chosen rule.
Thus the total computational overhead per block of data by the second model is 1953
calculations.
Model 3:
Computational overhead for a 9 character key
1st computation: 27 calculations, 2nd computation: 27 calculations, 3rd computation:
27*9 calculations, 4th computation: 27 calculations, 5th computation: 27 calculations, 6th
computation: 27 calculations, 7th computation: 27*27 calculations +27*27 calculations,
8th computation: 27 calculations, 9th computation: 81 calculations, 10th computation:
81 calculations, 11th computation: 81*k calculations (where k refers to number of
computations for a random pick up of values). Thus the total computational overhead per
character by the fourth model is 2025 +81k calculations.
Algorithm 2:
If we go by the algorithm, the number of computations required to generate the sequence
at different grid points by different time stamps depends on the steps of time stamp,
nonce considered. In the given problem for a time stamp 10 units & a Nonce of 21, the
total number of computation needed to generate 20 character block of cipher text is
around 500 computations.
Model 2:
Complexity of the Model by construction
If we go by the construction of the algorithm, there are 14 computations in the algorithm.
All the computations are product of the values of ternary vector. For the basins generated
computation, it is the square of the values of the ternary vector. If n is the number of
values of ternary vector considered, we can say that the complexity of model2 is O (n 2).
Complexity of the Model by Strength
In the given algorithm a matrix key is used. This matrix key is multiplied with ternary
vector. On the generated values a sign function is used to convert all positive values to 1,
negative values to –1 and zero to 0. A sequence is generated. Based on the similarity of
values this sequence is divided to basins, These basins with different values are mapped
to plain text to generate cipher text. Thus this algorithm provides different keys with
variable length. Thus this algorithm is supposed to be safe in real time environment. Thus
if we go by the security of the algorithm, the complexity is exponential in nature.
Model 3:
Complexity of the Model by construction
If we go by the construction of the algorithm, there are 11 computations in the algorithm.
All the computations are product of the values of ternary vector. For the basins generated
computation, it is the square of the values of the ternary vector. If n is the number of
values of ternary vector considered, we can say that the complexity of model3 is O (n 2).
Complexity of the Model by its strength
In the given algorithm a matrix key is used. This matrix key is multiplied with ternary
vector. On the generated values a sign function is used to convert all positive values to 1,
negative values to –1 and zero to 0. A sequence is generated. Based on the similarity of
values this sequence is divided to basins. Basins will form the number of characters of the
alphabet. A combination of basins are used which increases the number of characters of
the alphabet. Values from the basins are mapped to the values of plain text in a random
fashion to form cipher text. Thus if we go by the security of the algorithm, the complexity
is exponential in nature.
Algorithm 2:
Complexity of the model
By Construction: In the given model, the data value depends on a constant value which
is the key; it also depends on time stamp and a nonce value. By changing the time stamp
different values of the sequence which is to be mapped between plain text & cipher text
can be generated. By changing the nonce also, different values can be generated which
provides good security against crypto analysis. The complexity of the model by
construction is O(nt).
By its Strength: Since in the given model present data depends on the key, time
stamps, nonce & Past data, there is no means by which the key can be guessed rather than
trying all combinations of keys. Thus the complexity of the model can be considered
exponential in nature.
TABLE 6.2: Comparative study of Different Generated Models in terms of Security
analysis by construction and crypto Analytical Strength
The objective of linear crypto analysis is to find a effective linear equation of the
form P[x1….xn] ex or C[y1…….ym]=K[z1…..zl]. where P & C refers to plain and
cipher texts and K be the key. Once a proposed relation is determined, the procedure is to
compute the results of the left-hand side of the equation for a large number of plain text
& cipher text pairs. If the result is zero for more than half of the time, assume that
K[z1…zl] = 0 other wise it is assumed as 1. This gives us linear equations on key bits.
These equations can be solved to generate key bits.
Algorithm 1:
Model 1:
The model uses a sign function on the product of ternary vector and a matrix key
to generate the sequence. The sign function converts all positive values to 1, negative
values to -1, and zero with 0.This sequence is substituted for plain text to generate cipher
text. Thus it is impossible to generate the matrix key from the known plain text and
cipher texts. Thus this model is free from differential crypto analysis.
But this model uses a simple substitution technique to generate cipher text; it is
some what susceptible to linear crypto analysis. The Key can not be gained and not a
whole of information can be gained, but part of information can be gained in this model.
Model 2:
Since in this work, it uses a sign function to generate the sequence. Similarity of
values of the sequence will be used to generate basins which are used as variable length
keys to map plain text to cipher text. Since it uses a sign function, it is free from
differential crypto analysis and since variable length keys are used it is also free from
linear crypto analysis.
Model 3:
Since in this work, it uses a sign function to generate the sequence. Similarity of
values of the sequence will be used to generate basins. These basins are mapped to
characters of plain text. Then the values from basins are chosen randomly to generate
cipher text. Since it uses a sign function, it is free from differential crypto analysis and
since basins with different and variable values are mapped with characters of plain text in
a random way to form cipher text, it is also free from linear crypto analysis.
Algorithm 2:
The time stamp and nonce considered can be dynamic in nature, the number of rounds of
the algorithm and the number of values of the sub key generated is variable in nature.
These features provide sufficient security to the algorithm. Since the number of rounds is
variable in nature and a mod function is applied on the sub key values, it is relatively free
from differential crypto analysis. The Nonce value considered in the algorithm is
dynamic which gives variable number of sub key values which makes the algorithm free
from linear crypto analysis.
Model 1:
Model 2 12 3 Low Low Cipher text &
Known Plain
Text
Model 3 Indeterminate Indeterminate Low Low Cipher text,
Known Plain
Text, chosen
plain text &
cipher text.
2500
2000
1500
1000 Computational
Overhead
500
0
Model 1 Model 3
20
15
10
Changes in Key
0
Model 1 Model 2 Model 3 Algorithm 2
2.5
1.5
Changes in Plain text
1
0.5
0
Model 1 Model 2 Model 3 Algorithm 2
Table 6.5.1: Comparative study of the proposed Block Ciphers( Model 1 & Model2)
with standard algorithm like DES in terms of computational overhead, Data
overhead, complexity and security analysis.
Algorithm Computational Data Complexi Security Analysis
overhead per Overhead per ty by its
block of data block of data strength
DES * 49392 for a56 bit Equal Exponenti chosen text only
key al
Model 1 ** 486 for a 72 bit Equal Exponenti Cipher text only
key al
Model 2 1953 for a 72 bit Equal Exponenti Cipher text ,known
key al plain text only
70000 160
60000 140 Data overhead
50000 120 for a 20
100 character text
40000
Computational 80 40 character
30000 overhead per text
60
20000 character 40
10000 20 80 character
0 0 text
Model Model RC4 Model 3 RC4
3 4
APPLICATIONS
7. Applications
In applications like Mobile Adhoc Networks [39], Wireless Sensor Networks and
Broad Casting Applications, New symmetric encryption techniques and probabilistic
encryption techniques can well be used. It can also be used for Key encapsulating
mechanism (KEM) along with Public key cryptography, which can well be used in secure
electronic transactions.
A mobile ad-hoc network (MANET) is type of wireless ad-hoc network used for
mobile routers. The routers are freely movable and can change from place to place. Such
type of systems can be used independently or may be connected to internet. They can be
used at different places and at different times, usually with all nodes connected to few
hops of each other, and having data transfer at a constant rate. A new substitution
algorithms, or in particular a probabilistic encryption can well be used in MANET
environment to provide security to the constant data sent among different nodes of each
hop. The Unique characteristics of MANET’s pose a number of non trivial challenges to
security design
3. Limited Memory
Thus in this environments of MANET’s & WSN’s, there is a high possibility of pattern
matching to get any unauthorised information from the data being sent. To provide
security in this environment, the developed models can well be used. In probabilistic
model, multiple cipher texts will be generated for one plain text and any one cipher text
can be used for transmission. By the construction of the model, it is free from differential
and linear cryptoanalysis and also it is free from chosen cipher text attacks. In the second
algorithm, the encrypted data not only depends on the key but also on random nonce
value and a highly dynamic time stamp. Thus the probabilistic model and the time
stamped model avoids any statistical attacks and pattern matching attacks on the
encrypted data being sent, which makes the developed models suitable in these
environments. And more over the developed models require less computing power for
encryption & decryption process, they are well suited in these environments where the
data transfer is high.
Some more practical applications use hybrid encryption to deal with large
plaintext messages since the efficiency of the public key encryption algorithm is low and
computational head is high. As a main part of hybrid encryption schemes, Key
Encapsulation Mechanism [5,16] allows a sender to generate a random session key and
distribute it to recipient. In some communication scenario, the session key used in new
symmetric encryption model and as well probabilistic encryption models can be
encapsulated to designated group through their public keys. Anyone can encapsulate a
session key for a designated group and any recipient in the designated group can
decapsulate the session key with his private key. Thus the mechanism is secure against
adaptively chosen cipher text attacks in standard model, which provides ultimate security,
confidentiality and authentication to the transmitted data.
Chapter 8
8.1 Summary
This study represents the importance of Encryption of data for storage and
transmission. The significance of encrypted data can be identified in light of the
mushrooming applications and globalization of communication. The advantages of
encrypting data manifest themselves in the form of security & confidentiality in real time
applications. Encryption of data is of particular significance in applications like email, e
commerce, e-cash where highly vulnerable communication lines is accessed for
transmission of highly volatile data.
The study traces the development of various encryption models in a real time
environment in all their breath taking diversity and breakthroughs in Chapters 2. The
significance of the advances and adaptabilities is measured in terms of their diversity of
applications in myriad ways that we feel in our daily lives.
Chapter 3.2.2 builds a new variable length key block cipher algorithm. The
algorithm considers a matrix key and executes a sequence of steps which generates a
sequence. Based on the similarity of values this sequence is being divided into basins.
The basins with minimum values will be eliminated. Remaining each basin represents
one block of data. Depending on starting input plain text character, corresponding basin
is considered as a key. The procedure is repeated for certain plain text depending on
chosen value. Thus the cipher text obtained becomes very difficult to be broken without
knowing the key. The strength of the generated model is studied in terms of computing
power, Avalanche effect, and complexity of the model in terms of its construction &
strength. A Crypto analysis of the model is also identified to study the security of the
developed model.
Chapter 3.2.3 builds a new stream probabilistic cipher algorithm. The algorithm
considers a matrix key and executes a sequence of steps which generates a sequence.
Based on the similarity of values this sequence is being divided into basins. Each basin
represents one character. Each character in the plain text is replaced by a set of basins
based on chosen base value. Each basin value is replaced by random values from basins
to generate the cipher text. Thus multiple cipher texts will be developed for one plain
text. Any one cipher text can be used for data transmission. Thus the cipher text obtained
becomes computationally infeasible to be broken with out knowing the key. The strength
of the generated model is studied in terms of computing power, Avalanche effect, and
complexity of the model in terms of its construction & strength. A Crypto analysis of the
model is also identified to study the security of the developed model.
Topic 3.3 builds a model which considers not only the key, but also a time stamp
and a nonce value to generate sub keys which are mapped to plain text to generate cipher
text. The strength of the generated model is studied in terms of computing power,
Avalanche effect, and complexity of the model in terms of its construction & strength. A
Crypto analysis of the model is also identified to study the security of the developed
model.
A comparative study of all the developed models is discussed in chapter 6. All the
developed models are studied in terms of computing power, Avalanche effect and
complexity of the models in terms of their construction and strength. A crypto analytical
study of all developed models is also considered. Block Ciphers like Model 1 & Model2
of the first algorithm are compared with standard block cipher DES and ciphers like
Model 3 of the first algorithm & second algorithm are compared with stream cipher like
RC4.
8.2 Conclusions.
Encryption models presented in chapters 3.2.1, 3.2.2, 3.2.3 &3.3 ranged over new
block cipher techniques to stream cipher techniques. The first algorithm considers a
matrix key and executes a sequence of steps which generates a sequence. This sequence
is used to generate four different encryption models.
Model 2 is a variable length key block cipher technique. It is observed that, for a
given key, the total computational overhead of the model2 is 1953 calculations The
complexity of the model by its construction is O(n2), by its strength it is exponential in
nature. It is also observed that by slight variations in the key and the plaintext a lot of
variations in cipher text is identified which provides more strength to the generated
model. This algorithm is completely free from cipher text only and known plain text
attacks but may not be completely free from chosen plain text and cipher text attacks
where a part of information may be leaked out by crypto analysis.
A comparative study of all the developed models is presented in chapter 6. All the
models are compared in terms of computing power, avalanche effect, complexity of
models in terms of their construction and strength and a crypto analysis of the models is
also undertaken. The summary of the study is presented in table: 6.4. The models are
compared with standard algorithms like DES & RC4 algorithms and the results are
shown in Tables: 6.5.1 & 6.5.2.
The conclusions drawn above, reflecting the overall security and confidentiality
rates of transmitting data, confirm the improvement in the efficiency of transmitting data.
The methodology used in the dissertation can be used for evaluating new encryption
algorithms in terms of multiple parameters. Further, the quantitative data indicates
relationship between Random key considered, sub keys (Basins) generated,
computational power needed by the first algorithm and the strength and security of the
algorithm. It also identifies the importance of multiple parameters like keys, time stamps
and nonce values used in second algorithm in terms of its security & strength.
Another conclusion from the above study is freeness from public key attacks.
With probabilistic encryption algorithm (Model 3), a crypto analyst can no longer encrypt
random plain texts looking for correct cipher text. Since multiple cipher texts will be
developed for one plain text, even if he decrypts the message to plain text, he does not
know how far he had guessed the message correctly. Under this scheme, different cipher
texts will be formed for one plain text. Also the cipher text will always be larger than
plain text.
Chapter 9
FUTURE WORK
8. FUTURE WORK:
The present work deals with plain text being represented by numerical and
characters of English alphabet. The work can be improved so that it can support the
characters of not only English but also of other languages as well. The work can also be
improved to support not only text but also other forms of message transmission like
audio, video and images.
REFERENCES
2. Baocang Wang, Qianhong Wu, Yupu Hu: A Knapsack Based Probabilistic Encryption
Scheme, On Line March 2007, www.citeseer.ist.psu.edu.
5. Brics: Universally comparable notions of key exchange and secure channels, Lecture
Notes in Computer Science, Springer, Berlin, March 2004.
6. Sage.math.Washington.edu/home/jetchev/Public.html/docs/jetchev-talk.ppt- Broadcast
encryption schemes.
7. Brassard G.: Modern Cryptology , a tutorial lecture Notes on computer science , (325)
,(spring-verlas) .
8.Bruce Scheneier: Applied cryptography (John Wiley & sons (ASIA) Pvt. Ltd.
9. Carlone Fontaine & Fabien Galand: A Survey of Homomorphic Encryption for non
specialists, EURASIP Journal,Vol 07, Article 10.
10. Donavan G.Govan, Nathen Lewis: Using Trust for Key Distribution & Route
Selection in Wireless Sensor Networks, International Conference on Network Operations
& Management, IEEE Symposium 2008, PP 787-790.
14. Guo D, Cheng L.M., Cheng L.L: A New Symmetric Probabilistic Encryption Scheme
Based on Chaotic Attractors of Neural Networks, Applied Intelligence, Vol 10, No.1, Jan
99, pp 71-84.
15. Hamid Mirvazri, Kashmiran Jumari Mahamod Ismail, Zurina Mohd. Hanapi:
Message based Random Variable Length Key Encryption Algorithm, Journal of
Computer Science, pp 573-578, 2009.
16. Hianyi Hu, Gufen Znu, Guanning Xu: Secret Scheme for Online Banking based on
Secret key Encryption, Second International Workshop on Knowledge Discovery & Data
Mining, Jan 23-25 2009.
17. Henry Baker and Fred Piper: Cipher systems(North wood books, London 1982).
19. Kaiping Xue: Study of improved key Distribution Mechanisms based on two length
structure for wireless sensor networks, International conference on adv. Information
Technology, 2008.
20. Krishna A.V.N.: A new algorithm in network security, International Conference Proc.
Of CISTM-05, 24-26 July 2005, Gurgoan, India.
21. Krishna A.V.N., Vishnu Vardhan.B.: Utility and Analysis of some Encryption
algorithms in E learning environment, International Convention Proc. Of CALIBER
2006, 02-04 Feb. 2006, Gulbarga, India.
22. Krishna A.V.N., S.N.N.Pandit: A new Algorithm in Network Security for data
transmission, Acharya Nagarjuna International Journal of Mathematics and Information
Technology, Vol: 1, No. 2, 2004 pp97-108
23. Krishna A.V.N, S.N.N.Pandit, A.Vinaya Babu: A generalized scheme for data
encryption technique using a randomized matrix key, Journal of Discrete Mathematical
Sciences & Cryptography, Vol 10, No. 1, Feb 2007, pp73-81
24. Krishna A.V.N., A.Vinaya Babu: Web and Network Communication security
Algorithms, Journal on Software Engineering, Vol 1,No.1, July 06, pp12-14
25. Krishna A.V.N, A.Vinaya Babu: Pipeline Data Compression & Encryption
Techniques in e-learning environment, Journal of Theoretical and Applied Information
Technology, Vol 3, No.1, Jan 2007, pp37-43.
26. Krishna A.V.N, A.Vinaya Babu: A Modified Hill Cipher Algorithm for Encryption
of Data in Data Transmission, Georgian Electronic Scientific Journal: Computer Science
and Telecommunications 2007 !N0. 3(14) pp 78-83.
31. Pandit S.N.N (1963): Some quantitative combinatorial search problems. (Ph.D.
Thesis).
32.Pandit S.N.N (1961): A New matrix Calculus, J Soc., Ind. And Appl. Math. Pp
632-637.
33. Pci Yihting: A Temporal order Resolution algorithm in the multi server time stamp
service frame work, International Conference on Advanced Information Networking &
Applications, AINA 2005, Vol 2m 28-30 March, pp 445-448.
35. R.S.Thore & D.B.Talange: Security of internet to pager E-mail messages (Internet for
India-1997IEEE Hyderabad section) pp.89-94.
38. Suhas V. Patenkar Numerical Heat Transfer and Fluid Flow 11-75(1991).
40. wikipedia.org/wiki/Mobile-adhoc-network
41. wikipedia.org/wiki/sensor_network