When building optical networks for 2.5Gbit/s and 10Gbit/s DWDM networks, PMD was not of any major concern, other factors such as chromatic dispersion and attenuation where the limiting factors for NRZ optical transmitters. At 40Gbit/s and 100Gbit/s, PMD is a serious concern for NRZ transmitters and receivers.
During 2009 I measured PMD values on several optical fibre paths to determine realistic values of PMD for installed fibre systems. While the variation in PMD is quite small in buried fibre systems, aerial fibre is exposed to the wind and temperature changes. Hence the variation in PMD of aerial fibre is far greater than in buried fibre cables.
I presented a paper on the results of my testing at the CEF conference in 2009. The paper is available here: PMD 2009 Presentation
After this presentation I did succeed in recording diurnal changes in PMD in an older section of aerial fibre.
The graph below is of an aerial fibre pair I measured during the summer 2009 over 10 days. You can clearly see the PMD rising during the day and falling during the night for the 10 days.
My overall observation of this testing is that if you are using old fibre (manufactured before 2002), then your will have to compensate for PMD for any 40 or 100Gbit/s services.
With the adoption of DP-QPSK and coherent detection for 100Gbit/s DWDM solution, the limitations of PMD have been overcome with electronic compensation. New coherent transponders can tolerate over 30ps of PMD and even 100ps.
During 2009 I measured PMD values on several optical fibre paths to determine realistic values of PMD for installed fibre systems. While the variation in PMD is quite small in buried fibre systems, aerial fibre is exposed to the wind and temperature changes. Hence the variation in PMD of aerial fibre is far greater than in buried fibre cables.
I presented a paper on the results of my testing at the CEF conference in 2009. The paper is available here: PMD 2009 Presentation
After this presentation I did succeed in recording diurnal changes in PMD in an older section of aerial fibre.
The graph below is of an aerial fibre pair I measured during the summer 2009 over 10 days. You can clearly see the PMD rising during the day and falling during the night for the 10 days.
My overall observation of this testing is that if you are using old fibre (manufactured before 2002), then your will have to compensate for PMD for any 40 or 100Gbit/s services.
With the adoption of DP-QPSK and coherent detection for 100Gbit/s DWDM solution, the limitations of PMD have been overcome with electronic compensation. New coherent transponders can tolerate over 30ps of PMD and even 100ps.
Her is a link to some PMD documents that I have come across that maybe useful for anyone doing some research into PMD: goo.gl/JU6YHA
ReplyDeleteAlso here are some of the raw results which someone might find useful.
goo.gl/qK56UQ and goo.gl/aZgGK5