Computer Engineering and Applications ›› 2012, Vol. 48 ›› Issue (36): 58-63.

Previous Articles     Next Articles

Genetic algorithm-based approach to embedded software WCSD test

LI Xianjie, ZHOU Kuanjiu, WANG Jie, HOU Gang, CUI Kai   

  1. School of Software Technology, Dalian University of Technology, Dalian, Liaoning 116620, China
  • Online:2012-12-21 Published:2012-12-21

基于遗传算法的嵌入式软件WCSD检测方法

李显杰,周宽久,王  洁,侯  刚,崔  凯   

  1. 大连理工大学 软件学院,辽宁 大连 116620

Abstract: To lessen the measurement error of software’s Worst-Case-Stack Depth(WCSD) and determine the system's memory requirement in resource-constrained embedded systems, an interrupt schedule model is constructed and a dynamic approach of embedded software WCSD test based on Genetic Algorithm(GA) is proposed to guide hardware design and software development efficiently. In addition, an experiment is conducted on an embedded software digital simulation platform; its results show that the model and approach can give a more accurate WCSD value, and can be conducive to reduce the total memory cost and improve reliability of the embedded software.

Key words: software test, Worst-Case-Stack Depth(WCSD), State Transition Matrix(STM), interrupt schedule, Genetic Algorithm(GA)

摘要: 在资源受限的嵌入式系统中,为了降低嵌入式软件最大堆栈深度(Worst-Case-Stack Depth,WCSD)的检测误差,从而确定系统内存容量,通过详细分析堆栈使用原因和中断类型,建立中断调度模型,提出基于遗传算法的WCSD动态检测方法以更加准确地指导嵌入式硬件设计和软件开发。基于嵌入式软件全数字仿真平台完成实验,对该模型和方法加以验证。实验结果表明该方法可测得较准确的软件堆栈深度上限,有助于降低内存开销和提高系统的可信度。

关键词: 软件测试, 最大堆栈深度(WCSD), 状态变迁, 中断调度, 遗传算法(GA)