[802.3BA] proposals for May 803.3ba, part 1

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

All,

Please find files attached that provide supporting material and content proposals for clause 86 of P802.3ba regarding two inter-related subjects. One subject is a proposal to improve the TDP test methodology for SR4 and SR10. The other is a proposal to provide a means to screen SR4 and SR10 PMDs for extended reach (XR) capability using this improved TDP test methodology or the existing methodology.

The proposed change to the TDP test methodology is an improvement in multiple ways. It allows a more accurate assessment of the interaction of the dispersion of the fiber with the highly dynamic optical spectrum of multi-transverse mode lasers, a spectrum which is strongly affected by modulation. The present chromatic dispersion impairment is only a surrogate. It is simply a static contribution calculated into the reference receiver / filter combination. It does not capture the dynamic effects, nor does it account for the actual spectral shape of the particular laser under test. Furthermore, including an actual chromatic dispersion assessment into the test modifies the compliance space for the transceiver maker, permitting the maximal range of trade-offs between distortion, jitter, and dispersion. It therefore can result in increased yields.

The proposals on extended reach address and conform to feedback received in contributions from representatives of system vendors Cisco, Juniper and Brocade as documented in their contributions to the XR ad-hoc. The attached supporting presentation contains summaries of their main concerns and requirements. These also describe, point-by-point, how the proposal conforms to them. Given the constraints expressed in these contributions, a proposal that applies OM4 fiber in the manner suggested provides the solution that satisfies the most requirements. I have also incorporated into the proposals the feedback received at the March meeting after the contribution on the same subjects, including the request to offer text formatted exactly as envisioned within the document to aid the editors. In addition, feedback received from several members on earlier drafts of this material are incorporated. My thanks to those who provided comments.

You will see that the proposal differs from that of the March contribution in a couple aspects. First, the improved TDP test is not reserved for the XR annex, but proposed as a normative modification, as it offers the broadly appealing benefits of a more accurate test and potentially higher yield to the transceiver makers. Second, the multiple scenarios listed in the March contribution, delineated by 25m length increments, have been distilled to a single case for each of two possible outcomes of the TDP test modification proposal, i.e. acceptance or rejection. Each of these cases aligns with the TDP test fixture in that it requires no change to the bandwidth of reference receiver / filter combination to configure the fixture to screen for extended reach parts. Each simply requires adjusting the length of the test fiber in the fixture.

The XR annex proposal explicitly imposes no changes to the PMD specifications. It simply defines a screening method that will allow sorting the module production stream by performance to address the projected needs of the market.

For guidance, the subject of each of the files is:

kolesar_01draft5_0509.pdf = supporting presentation

kolesar_02draft1_0509.pdf = TDP test modification, proposed content for clause 86

kolesar_03draft2_0509.pdf = TDP test for extended reach, proposed content for informative annex 86A if kolesar_02draft2_0509 is accepted

kolesar_04draft1_0509.pdf = TDP test for extended reach, proposed content for informative annex 86A if kolesar_02draft2_0509 is not accepted

The last two files will be sent separately in part 2 so as not to exceed the reflector size limit.

Your review and feedback is kindly requested to assist in comment resolution during the May interim meeting.