DSPs simplify PMD Tolerance Testing of Coherent Receivers

New Ridge has been at the forefront in defining test methodologies and designing and manufacturing equipment to make PMD tolerance testing of fiber optical receivers quantitative, repeatable, and comparable. Our PMD sources have set the standard for the generation of PMD states to ease the often difficult and lengthy process of PMD testing.

Then coherent happened. And everything changed….

The DSPs, built inside the coherent dual polarization receivers, proved they could mitigate PMD better than any optical PMD compensator (including the ones I’ve built). But what is almost equally amazing is the validation and testing of PMD tolerance is also dramatically simplified.

Recently I was requested to present my new recommendations for PMD and SOP tolerance testing to the specification writing committee of the IEEE P802.3ct 100 Gb/s and 400 Gb/s over DWDM systems Task Force. A link to a presentation, similar to the one I presented to the task force, is available here. The presentation describes how to measure coherent receivers for (1) PMD tolerance and (2) immunity to SOP changes. My recommendation was to have 2 polarization scramblers sandwiching one PMD artifact having PMD = DGDmax.

simplified PMD and SOP tolerance testing set-up for coherent receivers

You ask, “I only need to measure 1 PMD state? That’s mind blowing!” Sure is. But the DSP cancels PMD by creating the time-reverse of the fiber’s PMD impulse response in its FIR filter. Therefore, as long as the FIR filter is capable of correcting the worst case impulse response element, it can correct all the PMD. Hence you only need to measure DGDmax.

This information was received as a bit of a shock and with apprehension as it drastically departed from the PMD tolerance testing methodology that I (and others) have been evangelizing for over a decade. But the old methodology of measuring many states of PMD, with numerous DGD and SOPMD values, is only necessary for direct detect receivers. I understand the skepticism of this new PMD test proposal since it replaces the extended test using an expensive PMD Source. And reduced it to a much simpler and faster test, costing much less to implement. Therefore we hope to link a more thorough mathematical justification of the PMD=DGDmax-only test to this blog when complete.

And lastly, we have recently discontinued building and selling new PMD Sources, as they are obsoleted by the coherent receiver technologies. We rather focus on solutions that actually help our customers, such as PMD artifacts with DGDmax. (Even so, we will continue to support, repair and calibrate our customers’ PMD sources for the foreseeable future.)

In the meantime, I invite you to look over my presentation on SOP and PMD testing of coherent receivers, and send me your feedback , or comment below.