[email protected] (Anne & Lynn Wheeler) writes:
> for some topic drift ... in hsdt there were some fiber-links with 10**-9
> bit-error-rate with 15/16s reed-solomon FEC ... which resulted in
> effective 10**-15 bit error rate ... approx. the same as ibm mainframe
> channels of the period.

aka this was just under 30yrs ago ... hsdt was fortunate to have an
engineer that had been one of reed's graduate students and had done lot
of the work on reed-solomon ... hsdt was also working with cyclotomics
up in berkeley that had done a lot of reed-solomon error correcting
work, including for the cdrom standard (i would joke that i could get
much better technology in $300 cdrom than i could get in a $10,000
computer high-speed modem) 

... a little drift, old reference
http://www.garlic.com/~lynn/2005n.html#email860414

past cyclotomics talk at stanford "The Impact of Error-control on
Systems Design"

even now, for gigabit link & 10**-15 bit error rate ... a packet being
dropped for uncorrected bit error rate would be every couple weeks ...
much more likely to have dropped packets because of congestion at
intermediate nodes.

and as mentioned ... hsdt was already doing dynamic adaptive rate-based
flow control.kfor congestion control/avoidance.

-- 
virtualization experience starting Jan1968, online at home since Mar1970

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to