Networks · Fundamentals

Participant- and message-oriented protocols

Message-oriented transmission offers flexibility, participant-oriented transmission enables targeted communication. CAN communicates by messages; higher-layer protocols such as CANopen add direct addressing of the participants.

Message-oriented data transmission

Here the focus is on the data packets — the messages — themselves. A message put on the bus by a participant is identified by a unique identifier of its own. Its destination is not defined: it is up to each participant to decide whether or not to accept a message that has been sent. A message may therefore be accepted by no participant, by one, or by several. Since the identifier generally also defines the priority for bus access, it is possible to give high-priority messages access with a short latency time, even under high bus load.

The producer normally does not know who consumes its message, which makes a confirmed exchange impossible. That is why message-oriented protocols generally apply the principle of error signaling rather than that of confirmation.

Message-oriented transmission therefore views the exchange as a process between systems or components without the identity of the participant being in the foreground. Each message contains all the information the consumer needs to process it. This approach allows a loose coupling between producer and consumer: changing one end of the communication chain does not necessarily require changes at the other end.

Participant-oriented data transmission

Conversely, participant-oriented transmission focuses on the identity of the communicating participants. Here it is crucial to know who sends and who receives. The exchange is organized around the participants and their relationships, and data is exchanged on the basis of addresses, node numbers or module identifiers.

The messages generally carry the destination and source addresses of the participants involved. Special addresses are reserved for transmission to a group of participants or to all of them (broadcasting). This means, among other things, that the messages transmitted to each participant must be explicitly specified.

And how does CAN communicate?

A CAN system generally communicates by messages. Each message carries a unique identifier that indicates its content and determines its priority; the consumer and producer, on the other hand, are not directly identified. Messages are therefore transmitted according to their importance and not according to the specific participant sending or receiving them. All the nodes in the network listen to all messages and independently decide, based on the identifier, whether a message is relevant to them. This allows for a great deal of network flexibility and scalability.

Using higher-layer, CAN-based protocols such as CANopen enhances this message-oriented communication and makes it possible to address participants directly. Each device in a CANopen network has a unique node ID, used in the messages to determine from which device a message was sent or for which device it is intended. Control commands and configuration data can thus be exchanged in a targeted manner, usually unicast, between a client and a server, following the request-response model.

In summary