The message Frames transmitted on a CAN bus do not contain the adresses of either the transmitting node or of any intended receiving node. This message-based approach eliminates data bits that would otherwise use some of the available bus bandwidth. It also simplifies the communication software, allows both node-to-node and broadcast messages to be transmitted simultaneously, and makes it possible to add new nodes to a system without updating other nodes. In lieu of transmit and receive node addresses, messages are labeled by an identifier (ID) assigned to one or more nodes on the network. All nodes receive the message and perform a filtering operation. That is, each node executes an acceptance test on the identifier to determine if the message- and thus its content- is relevent to the perticular node. Only the node(s) for which the message is relevant will process it. The identifier also contains data that specifies the priority of the message, and allows the hardware to arbitrate for the bus if several nodes attempt to transmit simultaneously. Every node on the bus validates every message. Corrupted messages trigger automatic retransmissions. Error detection and error signaling features and fault-confinement measures are defined in the CAN standard. They make the CAN bus very reliable, even in noisy environments, by ensuring that the information communicated between nodes is correct and cosistent. If a node is faulty, it will eventually go into a "Bus Off" mode that shuts down to prevent it from disturbing the traffic flow.