[e2e] Random losses on very high speed networks

[James P.G. Sterbenz [ssh]]

At 02:48 PM 7/21/2003 -0500, Saad Biaz wrote:
>Hi:
>
>
>Someone recently claimed on this list that on very high speed networks,
>random losses (unrelated to congestion) may well occur.

Well, since the discussion strayed to terminology, and no one has mentioned
this so far...

High speed optical links get FEC.  This allows service providers
to increase the spacing of amplifiers and regenerators.  So, it is conceivable
that if there are not sufficient network engineering margins, the link FEC
might not fully cover channel errors.  This would then result in the loss
of a frame (e.g. IP in GFP).

This would also be the case for O-CDMA (optical CDMA).


>Does anyone have some reference or numbers on the loss  rates?

Carrier folk ought to, but might not want to divulge a network engineering
parameter.  Don't know if this would be part of an SLA a customer would
see.

Cheers,
James Sterbenz
BBN Technologies