Security evaluation of RC4 using big data analytics

Abstract

Transport Layer Security (TLS) protocol aims to provide confidentiality and integrity of data across untrusted networks. TLS supports several encryption options, including the RC4 stream cipher. Recent years, there is a growing interest for the crypto researchers do cryptanalysis on RC4 algorithm, they provide a comprehensive view on the RC4 cryptanalysis. This paper presents a new method based on big data processing technique to analyze the security of RC4. We drive the powerful cluster to execute the analyzing program, which invokes the RC4 algorithm for producing the stream cipher derived from the inputted random key generated by locality-insensitive hash function with actually predictable number. With the manufactured stream cipher, the program inspects the statistical feature about the byte value distribution at each position and such feature among byte pairs. During the implementation, we parallelized our program such that it exploits the computational resources of the distributed computing system by using our own modified Mapreduce structure, then we can get the distribution derived from a huge amount of key.