Source Code Optimization and Proling of Energy Consumption in Embedded Systems

This paper presents a source code optimization methodology and a proling tool that have been developed to help design-ers in optimizing software performance and energy in embed-ded systems. Code optimizations are applied at three levels of abstraction: algorithmic, data and instruction-level. The pro ler exploits a cycle-accurate energy consumption simu-lator [3] to relate the embedded system energy consumption and performance to the source code. Thus, it can be used for analysis (i.e., to nd energy-critical sections of the code), and for validation (i.e., to assess the impact of each code op-timization). Code optimizations and proling tool are used to opti-mize and tune the implementation of an MPEG Layer III (MP3) audio decoder for the SmartBadge [2] portable em-bedded system. We show that using our methodology and tool we can quickly and easily redesign the MP3 audio de-coder software to run in real time with low energy consump-tion. Performance increase of 92 % and energy consumption decrease of 77 % (over the original executable specication) has been achieved for MP3 audio decoding on the Smart-Badge.