Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Jitter and PMDs

I had a question for the whole PMD group. 

Why can't we seperate the jitter, mask etc. specifications for WAN and LAN

It seems to me that we can just copy the SONET sepcs for the WAN phy. That
way, any vendor who makes Sonet complaint transceivers etc. can use them for
10gigE also. Also, the methods and equipment to test for Oc192 is pretty
straightforward. The added BIG advantage is that we can support Regens etc.
and allow people to go even longer distances (We don't need to specify how
much distance - this can remain a non-standard spec). This will immediately
kickstart the 10gigE WAN.

On the LAN side, we might need more discussions since we are not likely to
want regenerators there. However, I still believe we can borrow a lot from
Oc192. This might be simplifying things too much, but we could look at Oc192
components and look at their jitter generation, tolerances etc. in the
broadband sense. Sonet looks more at jitter vs. frequency. We can just
measure without any filters and get a feel for what we can do with present
technologies. Of course, in the LAN case we don't need to consider transfer

Rohit Mittal
ONI systems

Say Bye to Slow Internet!