[Sis-ams] AMS RID BNSC-46

[Scott Burleigh]

Our third residual RID is this one:

    Sending a separate "cell_spec" MPDU for each cell in the message
    space should be optimised to sending a single MPDU that contains a
    list of the registrars for the cells, i.e. concatenate the "cell_spec"s.

My proposed disposition is this:

    Not accepted.  The proposed change would complicate the protocol by
    adding another MPDU format; currently, the cell_spec MPDU can be
    used in both registrar registration and registrar location.  The
    reduction in network traffic would be insignificant, and processing
    the cell_spec MPDU would become appreciably more complex.

This is actually quite separate from the node registration protocol 
changes that we hammered out during the meeting: registrar registration 
is a procedure that occurs very rarely and will involve a tiny number of 
message exchanges unless the message space is truly mammoth, which I 
very much doubt will happen in an on-board environment.  I feel pretty 
strongly that this is a false optimization, and I would really like to 
set it aside.

Scott
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ccsds.org/pipermail/sis-ams/attachments/20071017/3ee8fd76/attachment.htm