Re: [6tisch] downstream traffic adaptation

[Tengfei Chang <[e-mail]>]

Yes, I have talked with all the authors and we agreed that the current
version is ready for WGLC!

Tengfei

On Mon, Oct 7, 2019 at 3:03 PM Pascal Thubert (pthubert) <[e-mail]>
wrote:

> Hello Tengfei and all
>
>
>
> I see that the 2 issues of IETF 105
>
>    - Rules for CellList
>    - Downward traffic adaptation
>
> Were discussed and there was no objection. Do you fell ready for WGLC now?
>
>
>
> All the best
>
>
>
> Pascal
>
>
>
> *From:* 6tisch <[e-mail]> *On Behalf Of *Tengfei Chang
> *Sent:* vendredi 4 octobre 2019 07:25
> *To:* Pascal Thubert (pthubert) <[e-mail]>
> *Cc:* 6tisch <[e-mail]>
> *Subject:* Re: [6tisch] downstream traffic adaptation
>
>
>
> Hi Pascal,
>
>
>
> Yes, it is implemented and tested, the result is shown in the (long) email
> "[6tisch] validating the downstream traffic adaptation in
> draft-ietf-6tisch-msf-06".
>
> I attached the figure here as well.
>
>
>
> Though that email is long, all the information is the explaination of the
> result figure, trying to make it clear.
>
> You can check the figure directly, I think is understandable for people
> following the msf well.
>
>
>
> Tengfei
>
>
>
> On Fri, Aug 23, 2019 at 4:23 PM Pascal Thubert (pthubert) <
> [e-mail]> wrote:
>
> Looks good Tengfei. Did you already implement and test?
>
>
>
> Pascal
>
>
>
> *From:* 6tisch <[e-mail]> *On Behalf Of *Tengfei Chang
> *Sent:* vendredi 23 août 2019 14:28
> *To:* 6tisch <[e-mail]>
> *Subject:* [6tisch] downstream traffic adaptation
>
>
>
> Hi All,
>
>
>
> As discussed in IETF 105, we are going to support adapting downstream
> traffic.
>
> And now this feature is added in draft-ietf-6tisch-msf-06 published one
> weeks ago.
>
>
>
> Some main changes related to this feature are:
>
>    1. After joined the network and got Rank, the node will only add one
>    Tx Cell to its parent (NOT add Rx Cell)
>    2. When adapting downstream traffic, the mote will check the cell
>    usage on the Rx Cells.
>    3. At each Rx cell to parent, the NUMCELLELAPSE increments
>    4. If there is any frames received on the Rx cell, the NUMCELLUSED
>    increments
>    5. If NUMCELLELAPSE reaches MAX_NUMCELLS, check the percentage of cell
>    usage, if higher than the high threshold, add one Rx cell to parent, if
>    lower than the low threshold, remove one Rx cell to parent.
>    6. The AutoRx cell is used for downstream as well and will never be
>    removed.
>
> Those changes are integrated into version 06 content.. You can read it to
> check exactly how we did the changes.
>
> Please let me know if you have comments on them.. Thanks!
>
>
>
> Tengfei
>
>
>
> --
>
> Chang Tengfei,
>
> Postdoctoral Research Engineer, Inria
>
>
>
>
> --
>
> ——————————————————————————————————————
>
>
>
> Dr. Tengfei, Chang
>
> Postdoctoral Research Engineer, Inria
>
>
>
> www.tcahng.org
>
> ——————————————————————————————————————
>


-- 
——————————————————————————————————————

Dr. Tengfei, Chang
Postdoctoral Research Engineer, Inria

www.tcahng.org
——————————————————————————————————————