Why does HDLC, in fact some of the (or all) link layer protocols use
sliding window mechanism? Sliding window provides ways to handles when
packets are lost, duplicated or are out of order. Isn't it the task of
Transport layer. TCP does that over IP which is unreliable.
How such problematic cases are handled in case of ethernet ?
Re: sliding window
"Ask" <email@example.com> writes:[color=blue]
> Why does HDLC, in fact some of the (or all) link layer protocols use
> sliding window mechanism?[/color]
The variation of HDLC used by PPP does not normally use
retransmit-based error correction -- see RFC 1662.
Are you referring to RFC 1663? If so, the point of that
rarely-implemented feature is to allow PPP to operate when there is a
very high underlying link error rate, and when the link itself has a
short delay compared with the transport layer (TCP) timers. See also
RFCs 3366 and 3819.
> Sliding window provides ways to handles when
> packets are lost, duplicated or are out of order. Isn't it the task of
> Transport layer. TCP does that over IP which is unreliable.[/color]
Indeed. That's why it's not normally done.
> How such problematic cases are handled in case of ethernet ?[/color]
Ethernet retransmits on late collision, when the transmitter can
detect that something has gone horribly wrong with the message.
Otherwise, it doesn't do link error control, either, just like PPP.
James Carlson, KISS Network <firstname.lastname@example.org>
Sun Microsystems / 1 Network Drive 71.232W Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757 42.496N Fax +1 781 442 1677