Re: [tsvwg] Interim meeting agenda for TSVWG April Interim Meetings

[Spencer Dawkins at IETF <[e-mail]>]

Hi, Ruediger,

On Thu, Apr 9, 2020 at 2:18 AM <[e-mail]> wrote:

> Hi Spencer,
>
>
>
> I tried to get in contact with 3GPP when LTE and the QCI were introduced.
>
Thanks for sharing your experience with this - it's WAY deeper than mine!

>
>    - QCIs are no markings. They are assigned to their connections  from
>    policy node to terminal.
>
> (I questioned "markings" when I wrote it - thanks for clarifying what I
was trying to say)

>
>    -
>    - There’s been a 9 standard QCI set which was mapped in no easy to
>    aggregate way to their DiffServ scheme (GSMA IR.34). I proposed changes.
>    - The standard QCIs are not binding. There’s also no binding subset or
>    minimum subset to implement.
>    - The mapping of standard QCIs to applications is not binding.
>    - There are many more private QCIs, which saw deployment at that time.
>    - LTE and likely 5G schedulers are proprietary.
>
>  I stopped working on that.
>
>  The best idea that I’ve heard was to introduce a DiffServ scheduler at
> the LTE(5G) sender and run the wireless section without these QCIs on a
> best effort base for all mobile terminals (the have a terminus for that
> mode of operating the wireless section). That would require a rather strong
> push from the mobile operators.
>
I know that you and David Black co-edited
https://datatracker.ietf.org/doc/rfc8100/, which was aimed at DiffServ for
operator interconnection, Do you think something like RFC 8100 would be
helpful?

(and I agree with your point about mobile operators needing to encounter
enough pain to be receptive to that kind of change)

>  I didn’t follow their work on 5G QoS.
>
Completely understandable :-)

And thanks again for this note.

Best,

Spencer