x
<<   <   Page 2 / 2
reese_lightening 12/5/2012 | 3:32:47 PM
re: Nortel Rolls On With 40-Gig Hold on gentlemen, let's keep emotions in check:

1) When something is sold should not lead to the direct conclusion that it is good.

2) Yes Lucent spent a lot of time publishing papers on nonlinear effects of pre-comp. In the special circumstances they were investigating they were right. If a carrier deploys in other circumstances, it could work well for them.

3) The subject of the thread is "DPQPSK to be de facto?" and the answer is clearly no because:
a) DPSK and ODB have the market lead.
b) There are many more companies that can do DPSK than companies that can do DPQPSK, which might lead to lower cost because of competition.
c) Nonlinear effects are very detrimental and performance CANNOT match DPSK by merely reducing input power, this was demonstrated by several authors, not just the New Jersey guys. For some distances, DPQPSK just won't work. At shorter distances it is great, but you are probably paying for "great" when you only need "good".
boozon 12/5/2012 | 3:32:45 PM
re: Nortel Rolls On With 40-Gig The problem is not the launched power into the DCMs, which can be kept to low levels (as the ALU team did in their experiment).
The problem has to do with the overall dispersion map, which doesn't agree with PM-QPSK.

Nonlinear effects depend very much on the link conditions: dispersion map + launched powers + channel spacing + mixing different line rates.
It would be good to know what NT has effectively deployed: is there a mix of line rates? is there a safe guard band between 10G and 40G lambdas? how big it is? are they deploying their 40G lambdas over a third party line with DCMs or are they removing the DCMs from the line and transporting all existing 10G services on 4x10G muxponders or with their 10G technology that doesn't need DCMs?

boozon 12/5/2012 | 3:32:43 PM
re: Nortel Rolls On With 40-Gig Thanks, NoCopper.
Given that the answers you provide correspond exactly to the ALU OFC paper experimental conditions where does the discrepancy between the results come from? Could you recommend some technical papers in order to compare with the ALU work?
NoCopper 12/5/2012 | 3:32:43 PM
re: Nortel Rolls On With 40-Gig Q) It would be good to know what NT has effectively deployed: is there a mix of line rates?
A) Yes, a mixture of 10G and 40G.

Q) Is there a safe guard band between 10G and 40G lambdas?
A) No safe guard required. 10G and 40G can run in the 50GHz grid side by side.

A) Are they deploying their 40G lambdas over a third party line with DCMs
Q) This has been tested with different vendors and works perfectly even on DCM compensated links.
NoCopper 12/5/2012 | 3:32:41 PM
re: Nortel Rolls On With 40-Gig I would recommend reading:

http://www.opticsexpress.org/a...

There is also an interesting OFC 2008 Paper from ATT and NT on "Performance of a 46-Gbps Dual-Polarization QPSK Transceiver"

The information on the 10G/40G interworking and third party equipment is based on personal experience and multiple real world deployments that do exactly that.
<<   <   Page 2 / 2
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE