Burrows-Wheeler Post-Transformation with Effective Clustering and Integer Coding Through Vectorization

This research is mainly based on lossless integer Compression. In this research one of the most renowned method called Burrows-Wheeler transform (BWT) has been used. This method is used to sort data reversibly. After this transformation, most of the repeated characters come together. The result of BWT has been passed through Run Length Encoder (RLE) module to get the Run characters, Run Length and Run Character frequencies. The Run characters and Run Character Frequencies have been passed through Move-to-Front (MTF) module to get MTF number and final MTF list. Then the MTF number and Run length, which are non-negative integers, have been sorted using counting sort module. This module sorts the numbers according to the Run characters. The result is non-negative integers which have been coded using interpolative coding method. This method is also known as Non-statistical and Non-parametric Coding. The data has been decompressed using the same modules as compression but in reverse order. It results lossless output. To implement vectorization big files are broken into chunks and passed through each module, creating new processes. So, it works in parallel. This method results speed boost exponentially with file size while keeping compression factor intact if not better.