Bitmask-based code compression methods for balancing power consumption and code size for hard real-time embedded systems

In recent years, code compression has been frequently investigated for embedded systems to reduce memory use and power consumption. Among various compression schemes, dictionary-based ones are applied for their good compression ratios and rapid decompression engines. Bitmask-based code compression, which was derived from the dictionary-based ones, has been proven to have a superior compression ratio and rapid decompression engine. In this paper, we adopt the bitmask-based scheme and replace some of its dictionary entries to achieve greatly reduced power consumption while maintaining a competitive compression ratio. For a cacheless architecture, we propose three basic styles of replacement, namely by-access-saving, by-frequency, and by-block. Another procedure, called by-alignment, is applied afterward to further improve power consumption. According to the experimental results, the by-block scheme with the by-alignment procedure achieves the best result. In the best case, an increase of 1.61% in compression ratio can result in a 43.75% reduction in power consumption ratio.