DMCA
Dependability Analysis of Fault-Tolerant Multiprocessor Architectures through Simulated Fault Injection (Chapter 5 and 6) (1993)
BibTeX
@MISC{Clark93dependabilityanalysis,
author = {Jeff Clark},
title = {Dependability Analysis of Fault-Tolerant Multiprocessor Architectures through Simulated Fault Injection (Chapter 5 and 6)},
year = {1993}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Introduction Computer systems achieve fault-tolerance primarily through redundancy. Multiple versions of a software routine can be executed to overcome implementation errors in the application code. Hardware can be replicated and operated in parallel or sequentially, as a series of spares, to survive logic faults. Redundant software is expensive to develop, and increases memory requirements and execution time. Redundant hardware is difficult to design, and adds to the cost, size, weight and power consumption of a machine. Many fault-tolerant applications, such as the control of fly-by-wire aircraft and deep space probes, have physical limitations on the amount of redundancy that can be incorporated into a system. Cost is always a consideration when adding redundancy to improve fault-tolerance. The level of redundancy needed is determined by dependability requirements and the nature of the faults and errors that can be expected to affect a system. The behavior of processors in th
Keyphrases
simulated fault injection fault-tolerant multiprocessor architecture dependability analysis many fault-tolerant application multiple version deep space probe physical limitation redundant hardware introduction computer system execution time logic fault power consumption dependability requirement memory requirement fly-by-wire aircraft redundant software implementation error application code software routine
