1. 6tisch Avatar 6tisch
  2. admin
  3. meetings

Wiki

Clone wiki

meetings / 131206_webex

View History

Minutes Webex 06 December 2013, 6TiSCH WG

Note: timestamps in PST.

Taking notes (using Etherpad)

  1. Xavi Vilajosana
  2. Pascal Thubert
  3. Thomas Watteyne

Present (alphabetically)

  1. Alaeddine Weslati
  2. Cedric Adjih
  3. Diego Dujovne
  4. Giuseppe Piro
  5. Kuor Hsin Chang
  6. Maria Rita Palattella
  7. Nicola Accettura
  8. Pascal Thubert
  9. Pat Kinney
  10. Patrick Wetterwald
  11. Pouria Zand
  12. Raghuram Sudhaakar
  13. Thomas Watteyne
  14. Xavi Vilajosana

Recording

Slides

Action Items

  • Pascal to published proposed milestones to ML.
  • Thomas to create new repository for 6top interface draft on bitbucket.
  • Thomas to continue discussion about need for 6LoWPAN context management for 6top on ML.

Agenda

Minutes

  • [08.04] Meeting starts
  • [08.05] Administrivia
    • Note well, IPR, etc.
    • IETF89 registration open. Hotel prices are very high, Pascal to ask IETF whether no negotiated prices.
    • [Thomas] cannot take notes in Etherpad due to connectivity issues.
    • Approval agenda

      No issues raised. Agenda approved.

    • Approval minutes last call

      No issues raised. Minutes approved.

  • 6TiSCH Milestones
    • Proposal
      • 12/2013 – WG to adopt 6TiSCH terminology
      • 12/2013 – WG to adopt IEEE802.15.4e TSCH overview
      • 12/2013 – WG to adopt 6TiSCH architecture
      • 12/2013 – WG to adopt 6TiSCH minimal configuration
      • 04/2014 – WG to adopt 6top draft(s)
      • 04/2014 – WG to adopt 6TiSCH data model for CoAP
      • 11/2014 – Submit first draft of 6TiSCH minimal configuration to the IESG
      • 11/2014 – Submit first draft of 6top draft(s) to the IESG
      • 11/2014 – Submit first draft of 6TiSCH data model for CoAP to the IESG
      • 12/2014 – Submit first draft of 6TiSCH terminology to the IESG
      • 12/2014 – Submit first draft of 6TiSCH architecture to the IESG
      • 12/2014 – Evaluate WG progress, propose new charter to the IESG
    • Discussion
      • milestones of the WG presented together with the charter.

        Action item: Pascal to published proposal to ML.

      • work for minimal + yang models are doable within this year.
      • expect to adopt the 6top and 6tisch data model after IETF in London
      • submit the drafts of this group to the IESG by november.
      • december 2014 terminology,etc..
      • december 2014 proposal to recharter to the IESG
      • [Xavi] 6top ready for IESG submissino by 11/2014 could be hard. Risks are:
        • mapping to 15.4e may take time
        • as we progress work on dynamic scheduling, 6top interface might evolve
  • Status WG docs
    • architecture changes
      • JP will probably add some intro text to upcoming PCE-related work
      • goal is to support recharter in december
      • [Thomas] Why the term TCP?
      • [Pascal] Because JP will probably highlight problems of PCEP supporting only TCP.
      • Proposed changes to architecture draft:
        • suboptimal operation depending on the schedule
        • 6top can modify bundle sizes but cannot indicate that this is a local decision, or not.
        • describe chunk allocation scheme
        • describe more-follows and need-more bits
    • 6top split
      • Qin/Xavi agreed to split doc in two

        Action item: Thomas to create new repository for 6top interface draft on bitbucket.

      • 6top interface
        • list of commands
        • mapping to YANG data model
      • 6top sublayer specification
        • monitoring process
        • 6top to 6top protocol
        • IE formats
    • 6LoWPAN context
      • [Thomas] How do we distribute the context? Do we want to use 6top to distribute it

        Action item: Thomas to continue discussion about need for 6LoWPAN context management for 6top on ML.

      • [Xavi] What can you place on a context?
      • [Thomas] very flexible
      • Use contexts to reduce number of some bytes in packets. Use contexts to compress groups of addresses.
      • [Thomas] As we are defining a set of elements in the YANG model, we could add the ability to install contexts.
      • [Kuor Hsin] will people use it?
      • [Thomas] yes, I have encountered it.
      • [Pascal] We expect a way to agree implicitly about the contexts because if not it can break. Having to distribute context information requires coordination.
  • [08.42] YANG example
  • [08.47] OTF draft
    • http://tools.ietf.org/html/draft-dujovne-6tisch-on-the-fly-00
    • https://bitbucket.org/6tisch/draft-dujovne-6tisch-on-the-fly
    • published today
    • Addressed ML comments on bundles, discuss allocation policies, which metrics are used.
    • Bundle size is over provisioned for peak policy vs. dynamic allocation policy
    • [Diego] 6top manages the bundle size. If 6top changes bundle size, what will happen with OTF.
    • [Pascal] 6top knows how to do it, but OTF decides when.
    • [Diego] Need to define terms more precisely (e.g. reserve which is somewhere in between allocation and schedule)
    • [Thomas] 6top is providing statistics and other layers make decisions, e.g. OTF makes decision.
    • [Thomas] I don't think we want 6top to take decisions. 6top only provides statistics and commands. OTF is the component which decides when to resize a bundle.
    • [Pascal] bundle is the BW between 2 routers. 6top does not make the decision.
  • [09.00] PRMA paper
    • http://www.ietf.org/mail-archive/web/6tisch/current/msg01598.html
    • http://www.ietf.org/mail-archive/web/6tisch/current/msg01612.html

      D.J. Goodman, R.A. Valenzuela, K.T. Gayliard, B. Ramamurthi. Packet Reservation Multiple Access for Local Wireless Communications. IEEE Trans. on Communication, vol. 37, no. 8, 1989

    • recommended by Prof. Adam Wolisz
    • [Thomas] core idea is simple: contention with arbitration by the AP. Stop talking means returning the slot to the "available" heap.
    • [Thomas] Nice through simplicity. Direct applicability to 6TiSCH has challenges:
      • expecting that the AP coordinates everything packet may not play well in multi-hop scenario
      • a node needs to overhear and thus stay awake all the time to get all the ACKs
    • Concepts are applicable to 6TiSCH. Piggybacking information in the ACKs is similar to using a "more bit". A parent grants its slots to its children it is very close ideas to what we do.
  • [09.06] AOB

    No other business raised.

  • [09.06] Meeting ends.

Updated