Image Processing Reference
In-Depth Information
The CRC and acknowledgment fields are placed after the data field. CRC sequence is encoded on
bits, which is followed by a CRC delimiter at the recessive value. he kind of CRC adopted in CAN
is particularly suitable to cover short frames (i.e., counting less than bits). he acknowledgment
field, instead, is made up of two bits, that is to say the ACK slot followed by the ACK delimiter. Both
of them are sent at the recessive level by the transmitter. he ACK slot, however, is overwritten with a
dominant value by each node that has received the frame correctly (i.e., no error was detected up to
the ACK field). It is worth noting that, in this way, the ACK slot is actually surrounded by two bits at
recessive level, that is to say the CRC and ACK delimiters. By means of the ACK bit, the transmitting
node is enabled to discover whether or not at least one node in the network has received the frame
correctly.
The frame is closed by the end of frame (EOF) field, made up of recessive bits, which notifies all
the nodes the end of an error-free transmission. In particular, the transmitting node assumes that the
frame has been exchanged correctly if no error is detected until the last bit of the EOF field, whilst in
thecaseofreceiverstheframeisvalidifnoerrorisdetecteduntilthesixthbitofEOF.
Different frames are interleaved by the intermission (IMS) gap, which consists of recessive bits
and effectively separates consecutive frames exchanged on the bus.
15.2.2.2 Remote Frames
Remote frames are very similar to data frames. The only difference is that they carry no data (i.e.,
thedataieldisnotpresentinthiscase).heyareusedtorequestthatagivenmessagebesenton
the network by a remote node. It is worth noting that the requesting node does not know who is the
producer of the relevant information. It is up to receivers to discover the one that has to reply.
Indeed, the DLC field in remote frames is not used by the CAN protocol. However, it should be
settothesamevalueasinthecorrespondingdataframe,soastocopewithsituationswhereseveral
nodes send remote requests with the same identifier at the same time (this is legal in a CAN network).
In that case, it is necessary for the different requests to be perfectly identical, so that they can overlap
totally in the case of a collision.
It should be noted that, because of the way the RTR bit is encoded, if a request is made for an
object at the same time as the transmission of that object is started by the producer, the contention
is resolved in favor of the data frame.
15.2.2.3 Error and Overload Frames
Error frames are used to notify nodes in the network that an error has occurred. hey consist of two
fields: error flag and error delimiter. There are two kinds of error flags: active error flags are made
up of dominant bits, while passive error flags consist of recessive bits. An active error flag vio-
lates the bit stuffing rules or the fixed-format parts of the frame which is currently being exchanged;
hence it enforces an error condition that is detected by all the stations connected to the network.
Each node, which detects an error condition, starts transmitting an error flag on its own. In this
way, there can be from to dominant bits on the bus, as a consequence of the transmission of an
error flag.
The error delimiter, instead, is made up of recessive bits. After the transmission of an error flag,
each node starts sending recessive bits and, at the same time, it monitors the bus level until a reces-
sive bit is detected. At this point the node sends more recessive bits, hence completing the error
delimiter.
Overload frames can be used by slower receivers to slow down operations on the network. This
is done by adding extra delays between consecutive data and/or remote frames. he overload frame
format is very similar to error frames. In particular, it is made up of an overload flag followed by an
overload delimiter. It is worth noting that today's CAN controllers are very fast, so overload frames
have become almost useless.
