proposed experiment, invitation for volunteers
Dear colleagues,
In considering equalization for multimode fiber links, a key concern has
been the time variance of the channel. Here "channel" should be
understood to be a composite effect of lasers, optical launch
arrangement, fiber, connectors, splices and the environment. Current
state of the art suggests that if the time variance of this channel is
slower than 1 MHz, it can be adaptively equalized. The portion faster
than 1 MHz will have to be treated as noise. What proportion of channel
variation falls in the "faster than 1 MHz" category? This was the
question asked by Exercise 2 posted on our web page earlier.
Oscar Agazzi and Tom Lenosky have diligently worked over the last few
days to write a response to Exercise 2. I thank them both. Their
response is a proposal to conduct an experiment. The document is posted
on our web page:
http://www.ieee802.org/3/ae/public/adhoc/equal/
My thanks also to Abhijit Phanse, Kishore Kota and Giorgio Giaretta for
useful comments. I expect the authors to revise this document once more,
after allowing time for more comments on this reflector.
May I recommend that we proceed as follows:
- Let's examine this document, improve on it, then adopt it as a test
procedure.
- We should have at least two volunteers to conduct this experiment. If
you wish to volunteer, please contact me before December 1, 2000.
- Meanwhile, I will work out with Mike Hackert the logistics of
obtaining the set of DMD-challenged fiber reels to run this experiment
on.
- If you volunteer, you will be expected to present the results of the
experiment on or before the P802.3ae March 2001 Plenary.
Regards,
Vipul
vipul.bhatt@xxxxxxxxxxx
(408)542-4113