Java Code Analyser for Estimating Embedded Software Efficiency

Abstract

Nowadays, the amount of software in embedded systems is increasing, mainly due its growing complexity and tight time-to-market constraints. Thus, tools to support the embedded software development process are indispensable, including tools able to offer some estimates of very important non functional requirements like performance and power. This paper presents a tool, which analyses bytecodes generated from embedded Java applications and produces several statistics and allows estimate properties such as performance and energy consumption. To demonstrate the analysis supported by the tool, a set of experiments was performed using applications commonly found in embedded systems.