Impact of Software Fault Tolerance to Fault Effects in OS-Driven RT Systems

keywords: Fault, effect, error, fault tolerance, real time, operating system, kernel, task, software, firmware
Real-time kernels are often utilized to simplify the design of embedded time/safety-critical applications. However, embedded systems are sensitive to transient and other faults, each of which can lead to various errors at various system levels and can potentially result in system failure. In existing works, the real-time kernels sensitivity to transient SEU errors was studied only with the goal to classify SEU impacts to system behavior. In this paper, the study is extended to permanent faults being injected into the persistent program (firmware) memory and into the volatile data memory. Moreover, it is studied there how the sensitivity is affected if combination of software fault-tolerant techniques is built into various levels of the system. On the experimental basis, it is shown that some fault impacts are difficult to be softened by common fault-tolerant techniques.
mathematics subject classification 2000: 68M15, 68N25
reference: Vol. 33, 2014, No. 4, pp. 757–782