Randomness Evaluation of Modified Lblock Cipher
This thesis present the analysis conducted towards one of the latest Lightweight Block Cipher; LBlock Cipher and its modified version; Modified LBlock Cipher – Design 1 and Design 2. The analysis conducted is a statistical analysis, which is to study the randomness properties of these three block ci...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This thesis present the analysis conducted towards one of the latest Lightweight Block Cipher; LBlock Cipher and its modified version; Modified LBlock Cipher – Design 1 and Design 2. The analysis conducted is a statistical analysis, which is to study the randomness properties of these three block cipher algorithms. Among the important requirement when designing a block cipher algorithm is that the algorithm can act as a random number generator. However, randomness analysis has never been performed before towards LBlock Cipher. The aim of performing modification towards LBlock Cipher is to enhance its randomness results. Two designs were constructed for modification of LBlock Cipher. Modification were made in Confusion Function, S of Round Function, F by replacing the eight 4 X 4 S-boxes. For Modified LBlock Cipher- Design 1 four 16 X 16 AES S-boxes are used, whereas for Modified LBlock Cipher – Design 2 four 16 X 16 different S-boxes which has same properties as AES S- box are used.During experimentation, this research project considers full rounds LBlock Cipher and Modified LBlock Cipher – Design 1 and Design 2 algorithms, which all three algorithms accepts a 64-bit plaintext, utilizes an 80-bit key, executes in 32 rounds and produces a 64-bit ciphertext. NIST Statistical Test Suite consisting of sixteen tests separated into two categories; Non-Parameterized Test Selection and Parameterized Test Selection was used to conduct testing and analysis, and the significance levels were set from 1% - 5%. Blocks of ciphertext were generated from these algorithms and were concatenated to construct a binary sequence. Nine different data categories were used to generate inputs (plaintext and key), with each having 100 samples. Comparison were made between statistical analysis results gathered for all three algorithms. It is shown that Modified LBlock Cipher – Design 2 produced better analysis results and has successfully enhance the randomness results of its original design. At 1% significance level, this design is 43% better compared to the original design, and at 2% - 5% significance levels, it is more than 50% better. |
|---|