AN IMPROVED STATISTICAL TEST SUITE FOR CRYPTOGRAPHIC ALGORITHMS

Abstract

Information Security is one of the biggest challenges of today’s world and without cryptography Information Security cannot meet the current challenges. In Information security, cryptography is practiced to accomplish numerous goals together with CIA triad e.g. confidentiality, integrity, and authentication of data we sent via internet. One of the most important properties of a cryptographic system is a proof of its security. There are multiple cryptographic algorithms and applications that can be used. Amongst them, a user needs one of high performance with low cost. Similarly, there are different techniques to evaluate  strength of ciphertext  and to assess strength of these encryption algorithms and applications. These Evaluation parameters are Encryption time, Decryption time, Memory used by cryptographic application to generate ciphertext, Avalanche effect and Entropy. Entropy, Randomness testing plays a vital role in cryptography. Due to severe limitation in existing NIST STS and other software used for randomness testing this paper has proposed a better an improved statistical test suite. All the codes are rewritten in python. This software is tested with the same test data that is provided in the NIST document and satisfactory results have been found. The codes have been put together in a software package that can be deployed on linux server. In the development process much help was derived from the C implementation of the NIST test suite by NIST as well as an earlier code of the NIST suite in r4nd0m by StuartGordonReid