Re: [5gangip] Launching 6G Talk -- Issue 1

[Alexandre Petrescu <[e-mail]>]

Le 27/05/2020 à 16:57, Behcet Sarikaya a écrit :
> Hi Alex,
> 
> 
> On Tue, May 26, 2020 at 8:56 AM Alexandre Petrescu 
> <[e-mail] <mailto:[e-mail]>> wrote:
> 
>     In trying to identify what is 6G, one might think about requirements.
> 
>     A bandwidth requirement would need to leap to maybe 3 times more than
>     the earlier generation.
> 
>     We would thus talk about 10Gbit/s bandwidth to the end user, over the
>     air, for 6G.
> 
>     The latency exhibited by a ping command - the RTT field - would be
>     required to fit in a nano-second range.
> 
> 
> The above are kind of non-IETF issues.
> 
> Maybe a requirement on AI (artificial intelligence) maybe more relevant?

Behcet,

AI is one of these fields of development very important happening since 
a few years now, and strangely not reflected at IETF, despite its close 
relationship to communications and the Internet.

I mean, we did see at IETF Cloud meetings, IoT meetings, and so on.  But 
nothing about AI.

AI might have to do a lot with Vehicular networks, and it is from a 
Vehicular networks magazine that a poster here extracted 6G references.

AI might be very much a core network issue, rather than an edge link to 
UE issue.  And I suspect 6G cellular networks would be 'flat', free of 
imposed structures such as gNB-GGSN links, etc.  In that case, one might 
think that an AI core network might be useful for 6G as much as for a 
wired line backbone network.  So probably 6G and AI might not be an 
aspect to ponder over.

AI techniques might be needed to develop more intelligent encoding 
schemes for over the air codecs such as to bring even more bandwidth. 
What SCDMA might be for 5G maybe AI-DMA might be new features to think 
of.  But that is AI applied to PHY layer.

So...

Alex

> 
> Behcet
> 
>     Alex
> 
> 
>     Le 25/05/2020 à 16:43, Behcet Sarikaya a écrit :
>      >
>      >
>      > On Wed, May 20, 2020 at 11:18 AM John Grant <[e-mail]
>     <mailto:[e-mail]>
>      > <mailto:[e-mail] <mailto:[e-mail]>>> wrote:
>      >
>      >     On 20/05/2020 14:55, Alexandre Petrescu wrote:
>      >      > that message was
>      >      > saying something like Wuhan got the virus because of the 5G
>      >     deployment
>      >      > there; I am not sure whether Wuhan 5G is at 3.5GHz,
>     because that
>      >     is the
>      >      > only differentiating aspect of 5G compared to 4G that
>     might impact
>      >      > chemistry/biology
>      >     In UK the fear is about 26GHz, and predates covid-19. ITU
>     doesn't help
>      >     by saying there is an absence of evidence of harm,
>     conspicuously not
>      >     saying there is evidence of absence.
>      >
>      >     [snip]
>      >
>      >      > In building a hypothetical list of answers to the question
>     of what is
>      >      > 6G, what are its distinctive characteristics, like huge
>     bandwidth,
>      >      > almost-zero latency, quasi-universal coverage,
>      >     Those were all supposed to be features of 5G, yet so far only
>     eMBB has
>      >     been delivered.
>      >
>      >
>      >
>      > Can we identify this as the Issue 1 as indicated on the subject
>     line?
>      > Let's continue to discuss it.
>      >
>      > Hopefully also over time we identify other issues of interest in
>     the future.
>      >
>      > Let's go!
>      >
>      > Behcet
>      >
>      >      > I would definitely add an
>      >      > answer telling that 6G uses IPv6 natively
>      >     Or maybe it'll use something completely different. Mobile
>     operators
>      >     think IPv4 has too much overhead, and IPv6 headers are
>     bigger; they
>      >     also
>      >     think fixed addresses are a security risk, so changing to
>     IPv6 doesn't
>      >     even eliminate NAT.
>      >
>      >
>      >     [snip]
>      >
>      >      > PPS: Administratively, one might think to rename this list to
>      >     something
>      >      > like "6GANGIP" or?  Renaming a list has been traditionally a
>      >     difficult
>      >      > task at IETF
>      >     Well, 3GPP still has 3G in its name, so sticking with 5G
>     should be fine.
>      >
>      >     --
>      >     John Grant
>      >     Nine Tiles, Cambridge, England
>      >     [phone] and [phone]
>      > http://www.ninetiles.com
>      >
>      >     _______________________________________________
>      >     5gangip mailing list
>      > [e-mail] <mailto:[e-mail]>
>     <mailto:[e-mail] <mailto:[e-mail]>>
>      > https://www.ietf.org/mailman/listinfo/5gangip
>      >
>      >
>      > _______________________________________________
>      > 5gangip mailing list
>      > [e-mail] <mailto:[e-mail]>
>      > https://www.ietf.org/mailman/listinfo/5gangip
>      >
> 
>     _______________________________________________
>     5gangip mailing list
>     [e-mail] <mailto:[e-mail]>
>     https://www.ietf.org/mailman/listinfo/5gangip
>