Skip to main content

Minutes IETF123: schc: Wed 09:30
minutes-123-schc-202507230930-00

Meeting Minutes Static Context Header Compression (schc) WG
Date and time 2025-07-23 09:30
Title Minutes IETF123: schc: Wed 09:30
State Active
Other versions markdown
Last updated 2025-07-23

minutes-123-schc-202507230930-00

Meeting Information

SCHC WG info

Where / When

Meeting Information

The SCHC session will take place during Wednesday Session II on Jul
23rd, 2025; that's from 11:30AM to 13:00AM local time (UTC). The meeting
will take place in room Castilla.

Agenda

[11:30] Administrivia [10min]

  • Presenter: Chairs
  • Note-Well, Scribes, Agenda Bashing
  • WG / drafts Status
  • IP protocol Number and Ethertype (Remove UDP port?)

[11:40] SCHC architecture [20min]

[12:00] Updating RFC 8824 [15min]

  • Associated draft: draft-ietf-schc-8824-update
  • Presenter name: Marco Tiloca
  • Objective of the discussion: updates in versions -04 and -05

[12:15] TIPTOP news [5min]

  • Presenter name: Alexander Pelov
  • Objective of the discussion: SCHC Applicability To IP in Deep Space

[12:20] SCHClets [10min]

[12:30] SID allocation [5min]

  • Presenter: Laurent Toutain
  • Discussion

[12:35] Coreconf Management [5min]

  • Presenter: Laurent Toutain
  • Discussion

[12:40] Static Context Header Compression Over 5G Networks [10min]

[12:50] SCHC Adaptation for Geonetworking Protocol [10min]

  • Presenter name: Manoj Gudi
  • Info

[xxxx] AOB [QS]

Minutes

[11:30] Administrivia [10min]

  • Presenter: Chairs
  • Note-Well, Scribes, Agenda Bashing
  • WG / drafts Status
  • IP protocol Number and Ethertype (Remove UDP port?)

Pascal doing the notewell
AP: interesting presentation about SCHC zero yesterday at ANRW, could be
presented at the end of the session if we have time.
PT presenting the document's advancement
A SCHC Mooc is available
Drop the request of a SCHC UDP port number, it is actually of no use
because ports numbers are negotiated.
RM: Couldn't be onsite, let's discuss on the ML
AP: Check with Carles about the use-case of UDP for 6LO-SCHC before
removing it
RM: will be more active in future on the mailing list.

Ana on the chat: TheUDP Ports will help to deploy the SCHC architecture
easily. That's what we pointed out in the draft-schc-6lo

Action point: confirm on mailing list with Bob and others about the next
steps

[11:40] SCHC architecture [20min]

PT presenting

PT: This document is about showing the WG actual work in progress on the
understanding and definition of the architecture. It's not about
completing the document first and then starting a work on it. The gist
of the terminology is solid, we are still working on the terminology.

SCHC WG is expanding what had been defined in the LPWAN WG. Many things
were implicit by then and must be defined for SCHC to be generalized.

PT (p2): a main thing is how SCHC interacts with the stack.

Example of how SCHC worked on LPWANs. All always goes through the SCHC
stack. Explains the need to have No-Compression for LPWANs and in this
case.
Example of how that would work on a generic stack - loop in the stack.
New way to insert SCHC in the network stack.

PT (p3): the stack understands to dispatch a packet to SCHC for SCHC
processing; SCHC will "rewrite" the packet what will be dispatched
again, to SCHC again or to the stack if no further SCHC processing is
intended.

During the side meeting, many discussions on the terminolgy. We won't do
that during the current meeting, the discussion will continue on the ML.

Notion of SCHC Packet/SCHC Datagram, and renaming the SCHC Headers
SCHC Header is not defined anywhere. All we had was a compressed header.
See how terms were defined in previous RFCs and see if we really need to
add new terms.
Introducing strata, a new type of header has appeared to indicate which
instance/end-point/set of rules must be used to decompress a given
packets. How do we call the "legacy" header and the stratum header?
Proposal: SCHC stratum header becomes SCHC Control Header/Control SCHC
header and SCHC Payload becomes SCHC Data Header
LT: Control is ok but data introducing confusion since we are
compressing an header.
PT: We're not discussing the terminology now

Introduction of Domain, as a set of end-points
The term seems to be appropriate but still has to be defined: end-points
that participate to an instance, stratum domain, administrative domain?

PT: Thanks to the work on 802.15.4, inside a domain, we could have a
stateless operation of SCHC. Btw, this really illustrates how the work
is done on the architecture. Some needs emerge from the work in the WG.

RM on the chat: Need some way to identify Admin Domain in actual
packets. I have discussed this in the past on the list. Minimally divide
the SCHC ID.

[12:00] Updating RFC 8824 [15min]

  • Associated draft: draft-ietf-schc-8824-update
  • Presenter name: Marco Tiloca
  • Objective of the discussion: updates in versions -04 and -05

MT presenting

MT (p2): Document to obsolete RFC 8824
MT (p3): mostly editorial changes from -3 to -4
MT (p4): clarified the use of the "tkl" function and defined the
"osc.piv" function...
MT (p5): Clarified requirements, FD must be ordered
MT (p6): editorial changes and move to SCHC github in -4 to -5
MT (p7): Revised compression of the OSCORE Option and clarified OSCORE
compression when using the group mode of Group OSCORE. No changes on the
SCHC C/D process.

MT (p8): Clarifications on compression of CoAP options, since different
approaches are possible (discussed in draft-universal-option)
MT (p9): new registry for SCHC Compression of CoAP options, FID of type
"coap.option(x).subfield"
MT (p10): open TODO: update "tkl" function to take care of extended
tokens
MT (p11): two open errata for RFC 8824. The first one can be approved.
The second one is correct in its essence but it would require
clarification updates in RFC 8724, so it gets too complicated. It's
practically better to reject the second erratum. Instead, RFC 8824 can:
i) add clarifying text specific to CoAP

AP: ETA for finished document?
MT: probably 2 or 3 more revisions, hopefully sometimes next year. As
some point, we need a close look at some text parts to be possibly moved
out to other documents in CoRE (or vice versa) in the interest of
avoiding normative references to drafts. Those text parts should better
be placed in documents that we estimate to be sent to the IESG latest.
AP: we should revise our milestones accordingly

[12:15] TIPTOP news [5min]

  • Presenter name: Alexander Pelov
  • Objective of the discussion: SCHC Applicability To IP in Deep Space

AP presenting SCHC applicability to IP in Deep Space (TIPTOP WG)
AP: thanks to Marc Blanchet for attending the session
AP (p2): pitch of interplanetary communications: long, variable RTTs;
~Mbps throughput, several percent of packet loss (depending on distance)

AP (p3): explains why SCHC is of interest in this use case
AP (p4): 2-10% gains possible with SCHC
AP (p5): presents next steps, namely work on a profile for SCHC for
TIPTOP and of possible, have support in the TIPTOP meeting

SC: already have SCHC-over-DTN draft in WG, should check applicability
AP: clarify DTN to be disruption-tolerant network

MB: only frames are travelling into space right now, not packets. The
frame error rate is 0 because of overprovisioning of resources. Account
for this.

[12:20] SCHClets [10min]

AP presenting

AP (p4): formal definition of SCHClet
AP (p7): SCHClet configuration provided in JSON
AP (p8/10): running code of compressor/decompressor in C
The main benefits of SCHClets is to allow to implement only some
functionalities of SCHC and not the whole RFC8724.

SC: good work, we face the issue that we just wanted to use a subset of
SCHC. Would like to collaborate.
PT: please do not call it a piece if SCHC

MG: can you clarify what "atomic means"?
AP: something that you can take out and which can live on its own. maybe
rather "autonomous"

JF: how do you ensure that this would be compatible with full SCHC
implementations?
AP: you would indicate in the config what subset of functions the
SCHClets are using

[12:30] SID allocation [5min]

  • Presenter: Laurent Toutain
  • Discussion

LT presenting.

LT (p2): range of 400 SID values have been assigned to the WG
LT (p3): shows how SID allocation impacts compression (length of SID
encoding)
LT (): Open questions. Do we allocate SID for CoAP options if we do not
use it or want to change it? SID cannot be changed and are forever.
A revision of RFC9363 is needed (YANG data model for several reasons.
Next steps: Double check the current document, Provide SID allocation to
IANA

PT: The document is not stable so is it possible to use an exerimental
range of SIDs?
LT: The answer is yes

[12:35] Coreconf Management [5min]

  • Presenter: Laurent Toutain
  • Discussion

LT presenting.

LT (p2): introduce management rules, which are compression rules to
compress management messages to modify the Set of Rules.
LT: complex process to modify the rules and keeping the SoR
synchronized. Proposed solution: rule modification is not permitted,
only rule addition.
LT (p6-7): presentation of Coreconf methods for installing new rules
(iPatch, RPCs)
LT: Error management: no time to explain
LT: Conclusion: work in progress, existing implementations for OpenSCHC
and microSCHC.

MG: question about TTL
LT: explains how guard period would help
AP: asks Marco to comment our use of Coreconf in the context of Rule
management
MT: Already present in several documents the WG should look at

[12:40] Static Context Header Compression Over 5G Networks [10min]

AM presenting SCHC over 5G Networks
AM (p3): architecture
AM (p4):
AM (p5): entities placing over 5G, introduction of a new entity, the
S-UPF (standard UPF implementing SCHC C/D)
AM (p6): SCHC signaling over 5G during the registration procedure. SCHC
capability is added in the EMM capability message
AM (p7): SCHC params: device/context identification reusing the PEI
(permanent Equipment Identifier)
AM (p8): example of compression of a UL transmission between UE and Data
Network

Poonam Yadav: question on placement
AM: no time left to answer, let's discuss directly

[12:50] SCHC Adaptation for Geonetworking Protocol [10min]

  • Presenter name: Manoj Gudi
  • Info

MG (p2): why and how applying SCHC to Geonetworking
MG (p3): periodic and frequent messages, can jam the network
MG (p4): install SCHC context on autonomous vehicules
MG (p5): presentation of the ETSI stack (simplified)
MG (p6): Applying SCHC to GNTW protocol
MG (p7): SCHC compression: 2 types of rules, conservative (generic) and
aggressive (more specific, less interoperable)
Compression example: send 15 bytes instead of
MG (p8): on-air time evaluation, 9-13% less
MG (p9): graphs
MG (p10): SCHC architectural issues to address platoons, necessitates
help from the WG
MG (p11): what's next? simulations are needed

[13:00] A Framework for Secure Autonomic IoT Device Management in Constrained Networks [QS]

Presenter: Maryam Hatami
Document: https://dl.acm.org/doi/10.1145/3744200.3744775

AP: quickly presenting the paper and inviting people to go speak with
the authors.

[xxxx] AOB [QS]