DICTIONARY-BASED CODE COMPRESSION TECHNIQUES USING BIT-MASKS FOR EMBEDDED SYSTEMS By SEOK-WON SEONG A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science DICTIONARY-BASED CODE COMPRESSION TECHNIQUES USING BIT-MASKS FOR EMBEDDED SYSTEMS By Seok-Won Seong May 2006 Chair: Prabhat Mishra Major Department: Computer and Information Science and Engineering Memory is one of the most restricted resource in the embedded system design. Code compression techniques have been proposed to address this issue by reducing the code size of application programs. Dictionary-based code compression techniques are popular as they offer both good compression ratio and fast decompression scheme. Recently proposed techniques improve standard dictionary-based compression by considering mismatches. We propose a bit-mask based code compression technique to aggressively minimize mismatches and improve compression efficiency. This thesis makes three important contributions: i) it proposes a compression-aware bit-mask selection procedure for creating more matching patterns, ii) it develops a dictionary selection technique to minimize the code size based on the selected bit-masks, and iii) it presents an efficient code compression algorithm using the mask selection and the dictionary selection procedures to improve compression ratio without introducing any decompression penalty. To demonstrate the usefulness of this approach, we used applications from various domains and compiled for a variety of architectures. Our technique outperforms the existing dictionary-based compression methods by an average of 15%, giving a compression ratio of 55% 65%.