Re: [EToSat] New Version Notification for draft-lhan-problems-requirements-satellite-net-02.txt

[Arashmid Akhavain <[e-mail]>]

In addition to standardising the communications between the UEs and the
mobile network, 3GPP defines every single aspect of the overlay network
from its required NFs to the specifications of all the protocols and
interfaces. Communications between RAN and mobile core is handled through a
specific set of 3GPP defined protocols which up to now assumed a stable and
reliable transport underlay. The data plane for example uses GPRS (GTP is
carried in UDP) to connect user plane functions (UPFs). I do believe that
there are routing and forwarding challenges when it comes to mega
constellations.

Having said that however, as per my comment earlier, I share Christian's
point of view. Satellite networks, at least for now, are close ecosystems.
Existing satellite providers employ their own proprietary mechanisms to
connect their satellites to one another. There is no
interoperability requirement between satellite operators so far. Well...at
least I haven't seen one yet. Without such requirements, there won't be any
need for standardisation.

Now, when I raised this concern earlier, our colleague, Raymond Burkholder
in his reply wrote











*"In the not too distant future (as in, it is being worked on right
now),some operators will be carrying the traffic of other operators.
ie,backbone operators in the sky will have a good foot print of
groundstations, and will be providing backbone operations to
satelliteoperators who don't have a large ground footprint (this is the sky
tosky to ground scenario, vs the broadband up-down scenario).There then
needs to be some sort of interoperability of the carrier vsthe end-user in
terms of how to get a endnode-satellite's dataencapsulated across a
carrier's satellite network down to the ground andthen over to the
subscriber's site in a real-time fashion."*

This operation/business model is interesting from the standardisation point
of view and something we should perhaps find out more about.

@ Raymond Burkholder: Do you have any pointers you can share?

Best regards,
Arashmid

On Fri, Feb 25, 2022 at 8:57 PM Christian Huitema <[e-mail]>
wrote:

> On 2/25/2022 4:23 PM, Meiling Chen wrote:
>
> > 3GPP has regarded satellite network as 5G NG-RAN infrastructure, Network
> Operators need and want to be able to use satellite networks to provide
> services, the owner of satellite network should use the technology provided
> by IETF instead of private technology which is used currently.
>
> 3GPP solves a specific problem: standardize the communication between
> mobile phone and wireless network so phones can work and roam on
> multiple networks. I don't think we are close to the point were
> satellite phones can be designed to work on multiple satellite networks.
> Of course if multiple satellite network operators and device make to the
> IETF and asked to develop such standards, the IETF would most probably
> hear them. But the satellite operators are certainly not doing that now.
>
> EToSat has been addressing a different but related problem: make sure
> that end-to-end protocols like TCP, TLS, HTTP or QUIC work well on paths
> that include satellite links or satellite networks. Up to date, that
> work has mostly addressed the geostationary satellite, dealing with long
> latency. Constellations of low orbit satellites do not suffer too much
> from the long latency problem, but may introduce other issues such as
> maybe variable bit rate or variable latency -- I say maybe because this
> is just a guess. If there are such issues, we may need to improve
> transport protocols or their implementation. But I have not heard of
> specific requests to address that so far.
>
> >   To develop, we need to change the stateof privatization. From the
> perspective of economy, satellite networks will be shared in the future. It
> is impossible for each Service Provider to send a large satellite network,
> sure you know that.  Therefore, the technology of satellite network should
> be dominated by IETF and 3GPP like the current Internet and wireless
> technology.
>
> On the contrary, each existing provider is using their own set of
> satellites, be it Inmarsat, Iridium, Globalstar, SpaceX or their future
> competitors. The IETF is not in a position to regulate the industry and
> change that practice.
>
> -- Christian Huitema
>
>
> _______________________________________________
> EToSat mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/etosat
>