Timing analysis of reliable real-time communication in CAN networks

The Controller Area Network (CAN) is a fieldbus network with real-time capabilities. It is generally considered that CAN guarantees atomic multicast properties, through its extensive error detection/signalling mechanisms. However, there are error situations where messages can be delivered in duplicate by some receivers or delivered only by a subset of the receivers, leading to inconsistencies in the supported applications. In order to prevent such inconsistencies, a set of atomic multicast protocols is proposed, taking advantage of CAN synchronous properties to minimise its run-time overhead. This paper presents such set of protocols, focusing on the timing analysis of the supported reliable real-time communication. It demonstrates that, in spite of the extra stack of protocols, the real-time capabilities of CAN are preserved, since the predictability of message transfer is guaranteed.

[1]  José Rufino,et al.  Fault-tolerant broadcasts in CAN , 1998, Digest of Papers. Twenty-Eighth Annual International Symposium on Fault-Tolerant Computing (Cat. No.98CB36224).

[2]  Alan Burns,et al.  Replica Determinism and Flexible Scheduling in Hard Real-Time Dependable Systems , 2000, IEEE Trans. Computers.

[3]  Eduardo Tovar,et al.  Integrating inaccessibility in response time analysis of CAN networks , 2000, 2000 IEEE International Workshop on Factory Communication Systems. Proceedings (Cat. No.00TH8531).

[4]  Flaviu Cristian,et al.  Atomic Broadcast: From Simple Message Diffusion to Byzantine Agreement , 1995, Inf. Comput..

[5]  Alan Burns,et al.  Calculating controller area network (can) message response times , 1995 .

[6]  Sam Toueg,et al.  Fault-tolerant broadcasts and related problems , 1993 .

[7]  Jörg Kaiser,et al.  Achieving Fault-Tolerant Ordered Broadcasts in CAN , 1999, EDCC.

[8]  Eduardo Tovar,et al.  Distributed Computer-Controlled Systems: the DEAR-COTS approach , 2000 .