Question on identifiers in extended CAN

Due to the limited maximum data packet size (and other reasons) many higher level protocols built on CAN use some of the bits in the identifier to describe aspects of the data. In addition many protocols reserve blocks of identifiers for specific purposes so that although not every combination is used it is still helpful to have plenty of bits in the identifier. Why do you want to know the % of unused identifiers - I have seen some pretty weird CAN metrics but that's a new one for me.

Michael Kellett

I'm with Michael on this in being completely puzzled why you think knowing such a number would help you in any way.

The only generic answer would be "However many the application can make good use of". It's easy to imagine a protocol that uses all 29 bits of the ID for useful purposes. In a prior project I worked on, we were at

24 bits, and could have made pretty good use of the remaining 5 in retrospect.

Think of a CAN network with a lot of nodes, multiple physical buses, gateways and whatnot. We spent 8+8 bits on identifying the sending and receiving node of a message, 4 on flow control, 2 to identify the physical buses, 1 to distinguish two generally different modes of addressing (physical vs. functional), and 1 was reserved. The remaining

5 bits would have served a good purpose to either implement a rolling frame sequence counter, or identify the command contained in the data part.

I would expect most protocols to be defined based on meanings of individual bits in the ID, not on counting different ID values.

Thank you Michael and Hans for replying.

I am a newbie in CAN and just wanted to clarify the necessity of 29 bit identifier .. fas i intend to come up with a information security model for CAN (with out any higher layer protocol) where in i have planned to split the 29 bit identifier in to multiple fields (as usually done in higher layer protocols). to justify my decision for allocating the number of bits for identifier i required that data ... if any one could even say how the project of adding a information security model to CAN sounds like...??