[e2e] Random losses on very high speed networks
cannara at attglobal.net
Mon Jul 21 17:29:40 PDT 2003
Saad, if you mean other than LAN radio (wireless LAN...), then microwave and
optical links can be subject to losses that occur because of variations in the
path. These are normally corrected (FEC, etc.), because many physical
features of such links aren't well controlled, as when birds can fly across a
microwave path, or when some fiber changes with heating. So, defenses are
designed in. Of course, correction codes have limits and some conditions will
indeed cause pkt loss. Switching of links, as when metropolitan fiber rings
are manipulated by add-drop multiplexors, will also create conditions that
might not be good, or foreseen. Finally, there's always device failure in any
of these systems, which will lead to losses. The bottom line is that the
quantity/duration of losses seen will depend on the thoroughness of the path's
design and provisioning, as well as its management and recovery systems, so
you'd have to survey appropriate providers of relevant network paths to get
estimates for typical values and ranges.
Saad Biaz wrote:
> Someone recently claimed on this list that on very high speed networks,
> random losses (unrelated to congestion) may well occur.
> Does anyone have some reference or numbers on the loss rates?
> Any hint will be highly appreciated.
> Saad Biaz
> Auburn University
More information about the end2end-interest