Data Protection in the CAN Network

Despite electromagnetically compatible design and differential signal transmission, in the harsh automotive environment it still cannot be assumed that no transmission errors will occur. The probability that corrupted CAN messages will occur undetected is extremely low. Responsible for this are the error detection mechanisms defined in the CAN protocol. On the receiver side, these include – in addition to the CRC (cyclic redundancy check) that can detect up to five errors within a CAN message – checking the format (form check) and the bit stuffing rule (stuff check). The sender performs bit monitoring and evaluates the ACK slot.

One bit time after finding a transmission error, the error-detecting CAN node interrupts regular data transmission by sending an error signal (error flag). This is made up of six dominant bits, which is an intentional violation of the bit stuffing rule, so that every CAN node in the network will perceive what until that point has been a local error, and therefore will likewise terminate the message transmission by sending an error flag. In the pursuit of data integrity, this method assures network-wide data consistency, which is very important.

Error handling is completed by the relevant sender repeating the aborted CAN message. The CAN bus must be available for this. This is the case if the relevant CAN node has detected 11 recessive bits. Due to priority-driven bus access, there is no guarantee of immediate repetition. In other words: The lower the priority of the given CAN message, the longer it may need to wait for repetition.