AN ALTERNATIVE S-BOX DESIGN METHOD BASED ON RANDOM SELECTION

Random selection based s-box designs have an important role in cryptology. There are many design proposals in this area. A new design method is proposed in this study. The proposed method has a different design architecture than the existing approaches found in the literature. The performance analysis of the proposed method shows that it may be an alternative to other methods.


Introduction
Cryptology is related with enabling two or more communicating parties to securely exchange information.The need for encryption and decryption has changed in new communication systems emerged with the industrial revolution.Old encryption systems are based on language characters and their conversions.However, in today's encryption systems, a set of 0's and 1's are needed instead of linguistic characters.In this scenario, the encryption and decryption procedure is based on a mathematical algorithm based on a secret key.Therefore, the encryption procedure can be defined by E (P, ke) → C function.Here ke is an element of key space.P clear texts, C encrypted texts.On the other hand, the D(C, kd) →P function is define decryption procedure.If kd = ke, the encryption system is a symmetric or secret key encryption system.On the other hand, if kd ≠ ke, the encryption system is called asymmetric or open key encryption system [1].
The main problem with symmetric cryptographic systems is that the sender and the receiver negotiate on a common key and prevent this key from being passed to third parties.Despite these disadvantages, symmetric encryption systems are faster than open key systems.Symmetric cryptographic systems are divided into block encryption systems and stream encryption systems [1].
Block cipher algorithms are one of the basic building blocks of modern cryptology.A robust block cipher algorithm should provide two key criteria for confidential communication -confusion and diffusion.In a block cipher algorithm, cryptographic structures known as substitution boxes (S-boxes) Many methods have been proposed in the literature for s-box design.One of the popular methods in recent years has become a random selection method.In particular, designs based on chaotic systems have become an active research area .Although there are many advantages, computer simulations of chaotic systems are difficult.A design method that can be an alternative to chaotic systems is proposed in this study.The performance measures of the proposed method are better than many chaos-based designs.
The rest of the study is organized as follows.In the second section, details of the proposed method are given.performance analysis of the proposed method in the third section.The results obtained in the last section are discussed.

Proposed Method
The process steps of the proposed method are given below.step 1.
Any text is selected.
The hash value of each word in the selected text is calculated using a hash function.

Performance Comparisons
Various measurements have been developed to make cryptologically good S-box designs.These  Linear Approximation

Conclusions
In this study, a method has been proposed to generate s-box designs using an entropy source that has good statistical properties.The outputs of hash functions are used as entropy source.In this way both statistical uniformity and safety requirements are ensured.It is seen that nonlinearity value is obtained better than the previous published 23 studies according to the analysis results.

Vol 7 ,
Number 2, 2017 European Journal of Technique EJT 230 are generally used to provide the confusion feature.In many modern block cipher algorithms, nonlinear single-element S-box structures [1].

metrics are briefly described:
Input output independence: Knowing input values does not change the unknown value of output values. Output input independence: Knowing some output values does not change the unknown value of input values.Output output independence: A partial information about the output bits does not change the unknownness of other unknown output bits. Non-linearity: This is the most important feature of S-box structures.It is a feature that prevents the S-box from being expressed in linear equations.These linear equations are used to decrypt the cryptographic systems in which the S-box is located.For this reason, S-box designs with high nonlinear characteristics should be used.Information integrity: Kam and Davida identified the information integrity as "each exit bit for each possible input value depends on all possible input values, not just the appropriate bits of the input bits."European Journal of Technique EJT 232 Avalanche Criteria: A measure designed by Fesitel for S-box structures and Substitution Permutation Network (SPN) based block ciphers.When an f boolean function changes one bit of the input bit, it tries to measure whether half of the output bits have changed. Inverse: This criterion is a desirable feature of S-box structures.If there is an individual mapping between the input and output values of an S-box structure, it can be reversed.If an Sbox structure can not be reversed, there are fewer output values than input values.In such a case, the output values are less unknown than the input values.

Table 1 .
Proposed s-box step 3. SHA3, the latest standard, is used as a hash function in the algorithm.step 4. The hash value is divided into 8-bit length groups step 5.If the decimal value corresponding to the 8-bit length value does not exist in the s-box, this value is added to the s-box.Otherwise, it is continued with a new 8-bit length value.step 6.These steps are repeated until the entire table is filled.Table 1 shows a sample s-box table run using this procedure.16x16 size s-box has been produced to make the comparison.But it is obvious that alternative sizes of s-box or mixing permutations can be produced in other dimensions.Vol 7, Number 2, 2017 European Journal of Technique EJT 231

Table 2 .
Table (LAT) and difference table (XOR table) are also used to show how resistant the additionally developed S-box structure is to linear and differential cryptanalysis.Table 2 compares the performance of the proposed method and random selection based algorithms using chaotic systems.Performance comparison