Transparent Data Compression and Decompression for Embedded Systems
Gustav Tano · KTH Publication Database DiVA (KTH Royal Institute of Technology) · 2008
Embedded systems are often developed for large volumes; the cost for storage becomes a critical issue in embedded systems design. In this report we study data compression techniques to reduce the size of ROM in embedded applications. We propose to compress RW data in ROM to be copied into RAM at the system initialization. The RW data will be collected and compressed during compiling and linking but before loading into the target platform. At the system initialization on the target platform, an application program will decompress the compressed data. We have implemented three compression algorithms and measured their performance with different configurations and data. Our experiments show that the Lempel-Ziv-Welch and Burrows-Wheeler Transform algorithms performs very well and are both suitable for data compression on embedded systems.