 | 1 | initial version |
My question is about what happens to these padding bytes when this frame is sent as 802.11 frame?
If the frame originates from an 802.11 interface, there are no padding bytes.
I think the reason for Ethernet padding was, originally, to make sure that a packet, when transmitted, will, once it's completely transmitted, result in a signal along all of a maximum-length Ethernet cable, so that, while it's being transmitted, if any other station on that cable tries to transmit, it'll either detect the carrier for the packet the other station is transmitting (that's the CS - "carrier sense" part of "CSMA-CD") or detect a collision between the the packet the other station is transmitting and the packet it's trying to transmit (that's the CD - "collision detect" part of "CSMA-CD").
That's no longer necessary on full-duplex switched Ethernet, but it's still used.
It's also not used for 802.11.
I'd have to look at various 802 standards to see if they indicate what should happen to the padding in an Ethernet frame if the frame is bridged from Ethernet to 802.11. If it's routed at the IP layer from Ethernet to 802.11, the router will probably strip off the padding.
