Re: [lp-wan] review of https://datatracker.ietf.org/doc/draft-moskowitz-lpwan-ipnumber/

[Robert Moskowitz <[e-mail]>]

On 7/7/22 05:45, Ana Minaburo wrote:
> Dear all,
>
> SCHC protocol number might become the identifier to know where the 
> compression has been used, either before L2 or in the upper layers. 
> This identifier will be helpful for IP tunneling or PPP.
>
> I found this subject interesting to widely use SCHC in different usages.
>
> Should we define a header for those protocols that do not have a next 
> header field? Or a general header containing a profile, expanded size, 
> a rule set id or other information?

SCHC is all about reducing transmitted data size.  At best we define 
rules for an implied next header.

In some uses, with SCHC as the IP Protocol, there is not even need to 
carry the SCHC rule ID, if this is the only rule between the peers.  
Otherwise, the SCHC RuleID SHOULD be the only data added by SCHC.

In my example of ESP{UDP{appmsg}}, I would apply diet-esp which would 
then compress UDP and maybe the appmsg (can do with MAVlink). In the 
UDP{DTLS{appmsg}} example, the UDP may get compressed to zero bytes, and 
with CID, Seq, and Length in DTLS, then compress the appmsg.

Something like that on a case-by-case basis.

So that is while I just have the SCHC assigned protocol ID and an 
optional RuleID.

>
> Ana
>
> On Wed, Jul 6, 2022 at 9:41 PM Robert Moskowitz 
> <[e-mail]> wrote:
>
>     Air to ground communications over a constrained wireless tech.
>
>     But that would just be the 'job' of the wireless tech provider!
>
>     But consider some avionics on the UA that is ethernet connected to
>     the controller that has the actual wireless adapter(s).  Then for
>     E2E, we need to do the SCHC above IP.
>
>     It is the use case in the draft?
>
>     On 7/6/22 13:13, Behcet Sarikaya wrote:
>>     Hi all,
>>
>>     I quickly read the draft.
>>     I don't understand why UA (unmanned aircraft, or a drone?) would
>>     need LPWAN?
>>
>>     Behcet
>>
>>     On Wed, Jul 6, 2022 at 8:38 AM Pascal Thubert (pthubert)
>>     <pthubert=[e-mail]> wrote:
>>
>>         Dear all
>>
>>         I reviewed
>>         https://datatracker.ietf.org/doc/draft-moskowitz-lpwan-ipnumber/
>>
>>         Please find my comments below.
>>
>>         In short, I believe that the document is useful and already
>>         well advanced and I would support the adoption when times
>>         come. But I need to ensure that a NH number is enough, vs. a
>>         new option that would transport SCHC parameters like Rule Set
>>         ID and/or Instance ID.
>>
>>         I suspect we'll make the call at IETF 114 but we need
>>         feedback by then, so please join in if you wish the draft to
>>         progress rapidly.
>>
>>         "
>>         If the Next Header in the IP header were SCHC, not ESP, a
>>         clear segregation of incoming traffic is directly supportable.
>>         "
>>         SCHC maintains P2P sessions (called Instances) that are
>>         associated with a P2P transport. Can we have one and one only
>>         SCHC Instance over SPI? How is the SCHC rule set determined?
>>         In the case of SCHC over PPP,  the PPP connection indicates
>>         the session one for one, and the rule set can be indicated
>>         as  URI in the IPv6-Compression-Protocol Configuration option
>>         see
>>         https://datatracker.ietf.org/doc/html/draft-thubert-intarea-schc-over-ppp-03#section-3
>>
>>         "
>>         Where it is possible with ESP's SPI to mitigate inbound
>>         packet processing challenges implicit SCHC would generate,
>>         DTLS header does not safely even provide this and a SCHC IP
>>         number is necessary to separate traffic.
>>         "
>>         Is the above undoable with TLS?
>>
>>         "
>>         Operation starts using Veriport's WiFi service.
>>         "
>>         Can we avoid brand names?
>>
>>         "IANA" section: The new protocol should be added to
>>         https://www.iana.org/assignments/ipv6-parameters/ipv6-parameters.xhtml
>>         shouldn't it? And then, by ricochet it will effectively end
>>         in https://www.iana.org/assignments/protocol-numbers.
>>
>>         Keep safe;
>>
>>         Pascal
>>
>>         _______________________________________________
>>         lp-wan mailing list
>>         [e-mail]
>>         https://www.ietf.org/mailman/listinfo/lp-wan
>>
>
>     _______________________________________________
>     lp-wan mailing list
>     [e-mail]
>     https://www.ietf.org/mailman/listinfo/lp-wan
>
>
> _______________________________________________
> lp-wan mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/lp-wan