I want to answer a past exams question with a clean/simple way about:
Why does ATM protocol use for virtual circuit numbering both VPI
and VCI logical numbers?
From what I've understood: The VCI changes through the nodes that it goes through and VPI has the destination path.
I found this also: The address is broken into a VCI and a VPI subfield.
Session following the same path for several nodes can be assigned the same VPI. Some switches only look at VPI, simplifying switch complexity.
Can someone help me get this cleared?
Best Answer
One reason--and I'm sure there are many, many more--is because you can have the same VCI on multiple different VPIs.
The VPI is 8 (OK, 8-12) bits, while the VCI is 16 bits. VPI 0 has VCI 35... but VPI 1 also has VCI 35, as does VPI 2. Guess what? VPI 3 also has a VCI 35. These VCIs don't necessarily interact.
This is probably more for the PVC (Permanent Virtual Circuit) implementation, common in ISP networks.
It's a bit like asking why you have a third AND a fourth octet in an IP address, or why you need a street AND city name in a postal address. ;)