Code Compression in ARM Embedded Systems Using Multiple Dictionaries

This paper presents a new code compression method where we use different dictionaries, all of them are based in traditional Huffman algorithm. Our method reduces code size by up to 34.5% (including all extra costs) and presents an improvement of almost 7% over the traditional Huffman method. We performed simulations using Simple Scalar and analyzes using the applications from benchmark MiBench and use the embedded processor ARM.