fgruman
commented
5 years ago Part of this goes back to History - originally, there wasn't an MC or NMC in the WDM model. There was only an OCH. The WDM cross connect was between OCH (MSA v1).
In MSAv2, we introduced the concept of MC and NMC. During the upgrade, the OCH --> NMC and MC is created to support the NMC. So the original proposal was to keep the cross connect on the NMC.
The cross connect on the ROADM does not represent the cross connect across a single WSS (in this case MC would make sense), but represents a logical connection between the external ports of a ROADM which include the degree WSS but also potentially the add/drop SRG complex. So the cross connect may go through not just one WSS, but many WSS and couplers/splitters.
There was also some questions about superchannel, but this has not yet been sufficiently discussed and agreed. Especially on the interface between the transponder and ROADM. The multi-lambda use case and specification has not been defined yet. So there may be a question if the interface between the transponder and ROADM would be a single port or multiple ports. This raised the potential of a N:1 cross connect (again not agreed).
So at this time, the cross connects are at the NMC. But get translated internally to the device into the MC connections through a specific WSS.
Why are ROADM cross connections in the OpenROADM model between NMCs? The use of the terms NMC and MC in OpenROADM seems to differ a bit from ITU-T definitions, e.g. in G.872, where a NMC is defined end-to-end as the concatenation of all MCs between source and destination.