Minutes IETF97: lpwan
minutes-97-lpwan-03

# Minutes, IETF 97 LPWAN WG Meeting #

Note: this document is formatted using Markdown
(https://daringfireball.net/projects/markdown/)

Agenda and Meeting information
==============================

    Meeting        :   IETF97 Monday, November 14, 2016 (KST)
    Time           :   15:50-17:50 Monday Afternoon session II (120min)
    Location       :   Grand Ballroom 2, Conrad Seoul
    Chairs         :   Pascal Thubert <pthubert@cisco.com>
                       Alexander Pelov <a@ackl.io>
    Responsible AD :   Suresh Krishnan <suresh.krishnan@ericsson.com>
    URLs           :   http://tools.ietf.org/wg/lpwan
                       https://datatracker.ietf.org/wg/lpwan/
                       https://www.ietf.org/mailman/listinfo/lp-wan
                       http://etherpad.tools.ietf.org:9000/p/notes-ietf-97-lpwan

* Intro and Status                                 [10 min]  (Chairs)
    * Note-Well, Blue Sheets, Scribes, Agenda Bashing
    * Charter presentation

* LPWAN Overview Document                          [80 min]  (Various)

    * LPWAN Overview Introduction [5 min] (Alexander, Pascal - stepping in for
    Stephen Farrell) * LPWAN Gap analysis [10 min] (Ana Minaburo, Carles Gomez)
    * Sigfox system description [20 min] (Juan Carlos Zuniga) * NB-IoT
    characteristics [20 min] (Antti Ratilainen - remote presentation) * WI-SUN
    overview [20 min] (Bob Heile) * LPWAN Overview Discussion [5 min]
    (Alexander, Pascal - stepping in for Stephen Farrell)

* LPWAN Static Context Header Compression (SCHC)   [30 min] (Laurent Toutain)

    * SCHC for IPv6 and UDP [20 min]
    * SCHC for CoAP [10 min]

* Optional (if time permits)                       [20 min]  (Various)

    * RoHC applicability in LPWAN [10 min] (Ana Minaburo)
    * LPWAN Fragmentation Header [10 min] (Carles Gomez)

Resources
=========

* Agenda: https://datatracker.ietf.org/meeting/97/agenda/lpwan
* Links to audio streams, meetecho and jabber:
https://tools.ietf.org/agenda/97/#97-mon-1550-lpwan * Presented slides:
https://datatracker.ietf.org/meeting/97/session/lpwan/

Summary
=======

WG meeting lasted 2 hours.

This was the first meeting of the group after its forming. The charter was
presented, with its two charter items. The rest of the meeting was dedicated to
presentations related to these two items.

Presenting draft-farrell-lpwan-overview-04 then
draft-zuniga-lpwan-sigfox-system-description-01 and
draft-ratilainen-lpwan-nb-iot-00

Draft farrell addresses the first chartered item - an LPWAN
characterization/overview document. It is a combination of four individual
submissions (the fifth will be submitted soon), one for each of the four
represented technology, and one on the gap analysis. After a quick intro on
that draft, 3 of the 4 baseline technologies were presented, Wi-SUN, SigFox and
NB-IoT (which was pushed to the end due to connectivity problems with
meetecho). The LoRa document presentation was postponed for the next IETF due
the LoRa Alliance meeting happening concurrently, one thing we'd like to avoid
in the future, and last minute cancellation of the editor. The main questions
on the LPWAN overview document were aimed at figuring out the level of detail
to be expected from the main draft, vs. the individual submissions. Questions
like - "What do we expect from the document? How much details (e.g. include
radio information)? Do we need the terminology/architecture from the Gap
analysis?" One sense was that the gap analysis portion may be removed before
shipping since it is of high by transient value.

There was a general consensus in the room that this is the way to go and that
we have everything necessary to advance with it. A call for adoption confirmed
this, with next step - calling on the mailing list.

Presenting draft-toutain-lpwan-ipv6-static-context-hc-00 and
draft-toutain-lpwan-coap-static-context-hc-00: The second charter item -
compression for IP/UDP/CoAP - was addressed by two documents in the second half
of the meeting. The first document focuses on introducing the general approach
and IP/UDP compression and has been around for more than 6 months. Interesting
discussion on the scope, the purpose, many people have read it, and there are
initial implementations. The call for adoption was accepted in the room, next
step - call for adoption on the mailing list. The extension document for CoAP
compression is a logical extension of the first document and the room also
decided to adopt the document along with the first one. The CoAP compression is
a much more recent one, with more need for polishing and technical work. The
chairs insisted that additional authors can/should contribute in case this
became a WG item, which was welcomed.

In short, very good support in the room for the three presented documents. A
little bit more than a third of the room has read the individual submission
drafts, and many are willing to contribute. Consensus in the room for adopting
all three documents, which should now be confirmed by the mailing list.

Action items
============

Call for adoption of LPWAN overview document on the ML.
Call for adoption of SCHC for IP/UDP on the ML.
Call for adoption of SCHC for CoAP on the ML.

Volunteers
==========

* Scribes
    * Dominique Barthel
    * Geraldine Texier
* Jabber
    * Carsten Bormann

Minutes
=======

* 15:50> Opening, agenda bashing, Charter presentation (10 min, Alexander,
Pascal)

* 16:00> LPWAN Overview Introduction (5 min)

    * Presenter: Alexander, Pascal (stepping in for Stephen Farrell)
    * https://datatracker.ietf.org/doc/draft-farrell-lpwan-overview/
    * Gathers description of 4 technologies and gap analysis.
    * Also atchitectural description and definition of common vocabulary across
    specific technologies. * 3 technology champion will present.

* 16:05> LPWAN Gap analysis (10 min)

    * Presenters: Ana Minaburo, Carles Gomez
    * https://datatracker.ietf.org/doc/draft-minaburo-lpwan-gap-analysis/
    * merged two prior drafts. Added content on RoHC
    * New content on benchmark changes, in particular scale issues, adaptation
    of the current mechanisms (timers, ...) * Architecture: need to decide how
    to name architecture components => need for terminology * Explains why
    neither RoHCv1 nor RoHCv2 are fit to the problem. * Juan Carlos Zuniga:
    what are the next steps? Ana: work with Stephen Farell to incorporate
    content into his draft. * Alexander Pelov (as chair): who as read the
    draft? Please pitch in on the mailing list about the terminology so we can
    move forward. * Suresh Krishnan: what is final form of the terminology
    info? * Suresh Krishnan: under which form? WG draft? Pascal Thubert: don't
    know yet. Current goal is to push Stephen's draft to RFC, including a
    terminology section.

* 16:15> Sigfox system description (15 min + 5mn Q&A)

    * Presenter: Juan Carlos Zuniga
    *
    https://datatracker.ietf.org/doc/draft-zuniga-lpwan-sigfox-system-description/
    * J-C presents the main characteristics of Sigfox. No roaming, no handoff.
    * Different bandwidth in FCC and ETSI to cope with FCC rules * Frames
    different format for uplink and downlink. * Payload 12 bytes up, 8 bytes
    down. * Keys are pre-provisioned. * 25 countries one opearator * Several
    current applications - different characteristics (e.g. packet sizes)
    depending on the application * Observing 6 down to 0 bytes payloads with
    current customers. * Pascal Thubert: what is the draft about? Juan-Carlos
    Zuniga: pretty much what I described here. Not sure it's long-lived. *
    Pascal Thubert: Sigfox does release much info, so this draft is of
    interest. * Juan-Carlos Zuniga: also info in LTN draft at ETSI. * Pascal
    Thubert: does the LTN doc intend to describe whole stack on top of which to
    run IP? * Juan-Carlos Zuniga: intention is yes indeed. If needed, we can
    fill the gap.

* Moved Antti due to meetecho voice issue, going to the WI-SUN presentation

* 16:25> WI-SUN overview (15 min + 5mn Q&A)

    * Presenter: Bob Heile
    * Field Area Network (FAN) by Wi-SUN Alliance.
    * Bob goes through use cases.
    * Based on 15.4g radio technology and 15.4e MAC enhancement (all included
    in 15.4-2015). * Mostly based on IEEE and IETF standards, with a little bit
    of TIA and ETSI. * High redundancy and frequency hopping => robutsness,
    rekliability and resilience to interference * PRotocol layers: Physical
    layer : FSK modulation, Data link layer, adaptation layer, networking layer
    * Wi-SUN Alliance sees significant value in work done here (at LPWAN) and
    strongly supports its objectives. * Wi-SUN already connects to Ethernet and
    WiFi trough router. Very interested in interoperability. * John Dowdel
    (Airbus Defence and Space): routing across networks, directly or through
    mediation device? Bob: through mediation device. * Pascal Thubert: plan to
    publish aggregated draft. Do you still plan to publish your own draft? Bob:
    yes, we do. * Alexander Pelov: so far, only star technology. Your
    contribution to help Stephen (on identifying the use-cases in which the
    LPWAN WG will be relevant to Wi-SUN) would be appreciated. * Carsten
    relaying Jabber room: is Wi-SUN technology "available"? Bob: yes, sign up
    as technology adopter (no charge) and you get access to it.

* Trying again to reach Antti

* 16:40> LPWAN Overview Discussion (5 min)

    * Presenter: Alexander, Pascal (stepping in for Stephen Farrell)
    * https://datatracker.ietf.org/doc/draft-farrell-lpwan-overview/
    * work on-going to get document focussed to the point.
    * Radio technology maybe has no direct implication, but may be educational
    for people coming from other areas in IETF.

    * Poll: who has red one of the drafts on the technologies - at least 1/3 of
    the room who thought the description of radio technology was useful?
    significant response. * A common terminology is needed - it will be
    integrated in this overview document, not on a seperate document * Last
    call is expected in april * Balance to be achieved between info in
    individual submissions describing technologies and this (WG intended) draft
    including info on candidate underlying technologies * Feedback from the
    room is needed on how far to go on the gap analysis document? How do you
    see the future of this document? * Juan-Carlos Zuniga: purpose of
    individual draft is also to explain interest from the technology
    vendor/alliance in the LWPA work. * Pascal Thubert: should we have a
    section on the common aspect or the common expectation, common for the
    technologies? * The question will be brought to the mailing list *
    Question: do we adopt this as a working group item? => room in favor =>
    mailing list call to have an adoption

* 17:00> LPWAN Static Context Header Compression (SCHC) for IPv6 and UDP (15
min + 5mn Q&A)

    * Presenter: Laurent Toutain
    *
    https://datatracker.ietf.org/doc/draft-toutain-lpwan-ipv6-static-context-hc/
    * Goal to compress all prtocol headers from IP up to CoAP. * -00 version
    due to multiple name changes, infact 3rd release. * star topology *
    Application running on the constrained nodes is known, star topology
    simplifies compression, no need to carry addresses * Flexible compression:
    depends on the application * Goal: compress IP, UDP ... in 0 byte *
    Juan-Carlos Zuniga: diagram on page 5, interface to the right is the same,
    not "Legacy" vs "IETF", right? * LT: a difference is that in the "legacy"
    we will not be able to do negociations for the compression * "Legacy"
    describes the whole horizontal block, not the interface to the right. * The
    application is known => the traffic is known so the fields can be
    classified as static, dynamic or computed * Compressor checks which rule
    applies to the packet at hand, and sends the rule number instead of the
    packet header (in some cases, along with parts of the header as well). *
    Decompressor rebuilds the packet header based on the rule number. * 3
    maching operators defined in the draft: ignore, equal or MSB (Most
    significant bit) * Dominique Barthel: packet "dropped" if no matching rule
    means do not pass through compression, not dropped by the network. LT:
    dropped at the compressor, but not necessarily transmitted. Best effort
    e.g. the packet may be too big and will effectively be dropped * Pascal
    Thubert: UDP checksum compression in RFC6282 passed IESG because frame has
    a 4-bytes MIC, much more powerful than checksum. In this case, we'll be
    asked what we have instead of checksum to achieve the purpose, so eliding
    checksum has to be optional and dependent on a better mechanism. *
    Formalism developped with YANG. YANG model for SCHC. * Laurent asking about
    adoption. Pascal Thubert: was going to ask at the end of the session, as WG
    co-chair. * Pascal Thubert: do you have running implementation on this ? *
    LT : one from Acklio, may be one from Columbia University * (?Rahul
    Jadhav?): dispatch info to distinguish between rule numer (i.e. compressed
    header) and regular legacy frame? * Laurent: could reserve rule number 0 as
    "uncompressed". * Consensus in the room for adopting the document => call
    on the mailing list

* (time not updated from original schedule, around 17:30) SCHC for CoAP (10 min)

    * Presenter: Laurent Toutain
    *
    https://datatracker.ietf.org/doc/draft-toutain-lpwan-coap-static-context-hc/
    * Compression will need have knowledge about what the thing is: server,
    client? So it knows what traffic/patternto expect and compress. Static
    context! * Propose new compression/decompression functions (CDF):
        * static mapping,
        * remapping (when your thing is a server)
        * entropy reduction

    * Juan-Carlos Zuniga: dependency between this and previous draft?
    Also,implementations? * Alexander Pelov: who has read this (relateviley
    recent) draft? a dozen hands. Who is willing ot contribute? same nuber. *
    Matthias Kovatsch (Siemens): this removes a lot of the flexibility in CoAP.
    question is ??? Laurent: can compress or not based on profile * Matthias
    Kovatsch : so, could compress HTTP2 as well? Laurent: yes. * Pascal
    Thubert: we will be calling for adoption for this document as well - both
    documents address one charter item - the idea of having them separated is
    good

* (time not updated from original schedule, around 17:40) NB-IoT
characteristics (15 min + 5mn Q&A)
    * Presenter: Antti Ratilainen (remote presentation)
    * https://datatracker.ietf.org/doc/draft-ratilainen-lpwan-nb-iot/

    * Antti Ratilainen: Addresses massive number of devices with low
    throughput. Max rate 30kbps and 60 kbps dowmnlink and uplink resp. *
    Operates on licensed bands. 3 deployment modes (substituting for one of the
    LTE carriers, in the LTE guradband, or standalone outside of an LTE
    spectrum). * data can be transported over Non Acces Stratum (normally meant
    to transport control data) that ends at MME, therefore not processed at
    eNodeB. * MTU is 1500 bytes.

* Pascal Thubert: Ask to read the 3 drafts in the next two weeks, respond to WG
adoption call on the mailing list.