Trying to understand how frames are structured between hosts and routers. My basic understanding is that a new frame is required to encapsulate a packet from each router.

Example being, if you have Host B and Host C, with 3 routers between them, and you are sending a packet from Host B to Host C.

This would mean that the packet would be packaged inside a Frame (say Frame B) at Host B, and then sent to the first router (i.e. Router R1). Router R1 would then unpack the packet, identify the need to go to Router R2 and encapsulate the packet inside an entirely new Frame (Frame R1).

This would continue until the packet completed the entire route, from Host B to Host C.

In total this would make a total of four (4) frames created to get the packet from start to end.

Is this generally correct? Please help


That is how it works.

Think about it this way: Host B is connected via ethernet, so it uses ethernet frames to R1. R1 is connected to R2 with PPP, so it uses PPP frames. R2 is connected to R3 with frame relay, so it uses frame relay frames. Etc.

All of the above layer-2 protocols have different, incompatible frames, so each hop needs a different frame to carry the same packet.

Also, understand that the layer-2 addressing is only valid on the LAN where the device is connected, so the Host B MAC address is only relevant on the LAN to which Host B is connected, even if the next router link is the same layer-2 protocol.

Even among IEEE related protocols (ethernet, Wi-Fi, token ring, etc.) that use the same 48-bit MAC addressing, the frames are different. If you have a LAN that has two different protocols, e.g. ethernet and Wi-Fi, you need a translating bridge that will keep the addressing, but it must translate one frame type to the other.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.