[Sls-slp] MC_OCF service and Master Channel Generation function in USLP

[John Pietras]

Members of the SLS-SLP WG ---
USLP has a Master Channel OCF service. Given it's name and its access point int the USLP stack (i.e., into the Master Channel Generation function), one might think that this service would inject OCF SDUs into every frame of a master channel, the same way the TM MC_OCF service is defined. However, that is not the case. For USLP, the MC_OCF service injects OCFs only for those VCs that are configured to carry OCF SDUs (i.e., the frames from some VCs in the MC will carry OCFs while frames from other VCs in the same MC will not). At first I thought that this service sounds more like what both the AOS and TM SDLPs call their VC_OCF services. However, after more careful reading I have come to the understanding that the USLP_MC_OCF service is neither the TM MC_OCF service nor the TM/AOS VC_OCF service:

  *   For the AOS/TM VC_OCF services, there are potentially as many sending and receiving users as there are VCs, where each service user has access to only the OCF SDUs carried by that user's VC;
  *   For the TM MC_OCF service, there are only one sending user and only one receiving user per MC, where each user has access to all of the OCF SDUs in the MC, and every VC in the MC carries OCF SDUs); and
  *   For the USLP MC_OCF service, there are only one sending user and only one receiving user per MC, where each user has access to all of the OCF SDUs in the MC, but only certain designated VCs actually carry OCF SDUs.

In general this comes through with a careful reading of the document, but there two places in the document that seem to contradict (or at least muddy) the definition of the service and initially sent me off track:

  1.  The last paragraph of 3.6.1 states "The user is identified with the GVCID of the VC. SDUs from different OCF users are multiplexed together for delivery within the OCF of the transmitted VC." If there is only one service user per MC, wouldn't the user be identified with the MCID of the MC? The second sentence is even more confusing - it make is sound as though there can be multiple sources ("OCF users") of SDUs that will all be fed into a single VC - seemingly the exact opposite of what I think the service is supposed to do. I'm not sure what this second sentence is intending to say.
  2.  4.3.8.2 states "If there is a user of the USLP_MC_OCF Service for a particular VC, an OCF_SDU shall be extracted from the OCF of the Transfer Frames and delivered to the user." "A user ... for a particular VC" could be (mis)interpreted to mean that there can be a different MC_OCF receiving user per VC in the MC, instead of (the correct) only one user per MC. Again, I don't know what this sentence is intending to convey.

Finally, there is an error in figure 4-14: the VC Multiplexing function should send its MC Frame to the underlying Master Channel Generation function, not the Master Channel Multiplexing function.

Figure 4-14 should be fixed (perhaps  - if it's not too late - along with the changes regarding repeated/systematic retransmission), and I also suggest that the above ambiguous statements be clarified or removed.

Best regards,
John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/sls-slp/attachments/20200128/37e42db0/attachment.htm>