Some authors (i.e. Tanenbaum, Stallings) reference ARQ protocols as layer 2 mechanisms for handling errors.

Even this page from Wikipedia says that

these protocols reside in the Data Link or Transport Layers

Which ARQ techniques/protocols are used for handling errors in Ethernet?

Or, different question... which L2 protocols use "Go-back-n", "sliding window" or "stop-and-wait" as ARQ?

1 Answer 1


Ethernet's link layer uses CRC32 in the Frame Check Sequence field, trailing the payload data. If FCS fails the frame is dropped. Other reasons for dropping a frame include link congestion, QoS, ACL filtering. In any case, there's no recovery and it's up to higher level protocols to cope with it (or not).

10 gigabit and faster speed PHYs may use additional forward error correction (FEC). Minor errors are corrected on the fly while larger errors get the frame dropped as well.

Most commonly, errors are handled by the almost ubiquitous TCP transport layer protocol. TCP uses a sliding send window that resends packets that have not been acknowledged within the usual round-trip time.

  • Do you know which L2 protocols use "Go-back-n", "sliding window" or "stop-and-wait" as ARQ? When you say "It's up to higher level protocols to cope with that" do you mean LLC?
    – T.G.
    Feb 12, 2018 at 21:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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