Communication Integrity in Critical Embedded Systems

We present, in this paper, a fault-tolerant approach to cope with accidental communication data corruption in critical embedded systems. One of the classical integrity approaches is the redundancy-based approach that consists particularly in replicating the message and sending all copies via the same communication channel consecutively or sending them via replicated communication channels. Yet, such approach is vulnerable to some cases of Common-Mode Failure.

So, we propose to diversify the copies to be sent via two independent proposals: i) diversifying either the error detection function (which generate the check bits) or ii) the data payload. This paper focus on the first proposal by presenting experiments and results to validate its effectiveness. Besides, it describes basic theoretical concepts of the second proposal. Our case study is the Flight Control System (FCS). Yet, our approach could be deployed in other systems for which we describe the key properties.