Coverage and the Use of Cyclic Redundancy Codes in Ultra-Dependable Systems

Abstract

A Cyclic Redundancy Code (CRC), when used properly, can be an effective and relatively inexpensive method to detect data corruption across communication channels. However, some systems use CRCs in ways that violate common assumptions made in analyzing CRC effectiveness, resulting in an overly optimistic prediction of system dependability. CRCs detect errors with some finite probability, which depends on factors including the strength of the particular code used, the bit-error rate, and the message length being checked. Common assumptions also include a passive network inter-stage, explicit data words, memoryless channels, and random independent symbol errors. In this paper we identify some examples of CRC usage that compromise ultra-dependable system design goals, and recommend alternate ways to improve system dependability via architectural approaches rather than error detection coding approaches.

Insights into the Sensitivity of the BRAIN (Braided Ring Availability Integrity Network)--On Platform Robustness in Extended Operation

“…Such faults may be common for implementations deploying 32-bit-based computing architectures handling frames. In [22], Paulitsch et al argue that the error detection coverage credit may even be less.…”

A Multi-function Error Detection Policy to Enhance Communication Integrity in Critical Embedded Systems

“…In fact, paper [2] details assumptions taken explicitly or implicitly by most of existing works and shows that they do not reflect sometimes the real system requirements. Unrealistic assumptions are likely to result in an inefficient policy.…”

“…Since the error detection policy depends closely on the problem assumptions [2], we detail here our specific assumptions which are: 1) our policy is implemented in the application layer so it is independent of the network architecture, 2) a calculation refresh cycle of some ms (10 ms for FCS), 3) no redundancy of the communication channels and 4) communication channels are i) memoryless (output at the time t depends only on the input at the time t and not on previous ones); ii) binary (two symbols 0 and 1) and iii) symmetric (the error probability is independent of the transmitted symbol).…”

Interactive Consistency on a Time-Triggered Real-Time Control Network

“…It is envisaged that even more stringent reliability requirements may be necessary for automotive safety-critical systems. This is as a result of the increasing number of automotive vehicles, and thus the increased number of hours of operation, as opposed to the aerospace applications, such as fly-by-wire systems in commercial aircraft [10].…”

“…Such ultra-reliable systems are required to have a failure rate of failures per hour or less [7], [10], [11]. Such systems are also required to have a maximum system service outage time (i.e., the maximum time that a specific system service, such as steering, is unavailable) of between 10 and 100 ms, depending on the application [7], [10].…”