• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>IEEE transactions on industrial informatics >Software-Based Control Flow Checking Against Transient Faults in Industrial Environments
【24h】

Software-Based Control Flow Checking Against Transient Faults in Industrial Environments

机译：工业环境中基于软件的基于控制流的瞬态故障检查

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Mechatronic systems operating in industrial environments are subject to a variety of threats because of harsh conditions. Industrial systems usually use commercial off-the shelf (COTS) equipment which are not robust and safe against hostile conditions and therefore require fault-tolerance considerations. This paper presents a novel and efficient method for online detection of control flow errors, called software-based control flow checking (SCFC). It is implemented purely in software and does not manipulate the hardware architecture of the system. Redundant instructions and signatures are embedded into the program at compile time and are utilized for control flow checking at run time. The signatures of the basic blocks are derived from the program graph. It is shown in the paper that SCFC method can increase single detection capability to 14.7% and the fault coverage to 6.12% averagely in comparison with other methods without any increase in memory and performance overheads. In the paper, besides experimental evaluations, analytical evaluations are also carried out, based on probability principles. The detection ability of each method used is thus computed. These computations verify the experimental results and show that SCFC can detect more errors than other methods suggested in literature. Considering the memory limitations in some (such as space) applications and the trend towards the requirement for faster execution of programs, we suggest a novel metric called fitness parameter which incorporates these. It is a better measure than the previously proposed ones since it considers the fault coverage, the memory overhead and the execution time (performance overhead) of each method simultaneously, as well as the detection capability.
机译：由于恶劣的条件，在工业环境中运行的机电系统会受到各种威胁。工业系统通常使用商用现成的（COTS）设备，该设备不坚固且不能在恶劣条件下安全，因此需要考虑容错性。本文提出了一种新颖，有效的在线检测控制流错误的方法，称为基于软件的控制流检查（SCFC）。它完全由软件实现，并且不操纵系统的硬件体系结构。冗余指令和签名在编译时嵌入程序中，并在运行时用于控制流检查。基本块的签名来自程序图。本文表明，与其他方法相比，SCFC方法可以将单个检测能力平均提高到14.7％，故障覆盖率平均提高到6.12％，而不会增加内存和性能开销。在本文中，除了实验评估之外，还基于概率原理进行了分析评估。由此计算出所使用的每种方法的检测能力。这些计算验证了实验结果，并表明SCFC可以比文献中建议的其他方法检测到更多的错误。考虑到某些（例如空间）应用程序中的内存限制以及对更快执行程序的要求的趋势，我们建议采用一种称为适应性参数的新颖度量标准，将这些特征纳入其中。它比先前提出的方法更好，因为它同时考虑了每种方法的故障覆盖率，内存开销和执行时间（性能开销）以及检测能力。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号