Hello 6TiSCH,

As agreed at the IETF 106 meeting, I start this thread to support the
continuation of the 6TiSCH WG and the work on this document.

https://tools.ietf.org/html/draft-tiloca-6tisch-robust-scheduling

The draft describes how 6TiSCH nodes can efficiently and pseudo-randomly
alter their original schedule at each slotframe. The resulting schedule
is robust and not vulnerable to selective jamming attack, while
remaining consistent as well as collision-free. Appendix A shows a
step-by-step example output from the proof-of-concept implementation at [1].

After the presentation at IETF 104, the draft was well received, and got
three positive reviews from Michael [2], Tengfei [3] and Pascal [4].
Reviewers' comments were addressed in the latest v -02 presented at IETF
105.

An open point concerns the renewal of keying material to permute the
schedule. The current draft discusses how such keys can be distributed
to a pledge by relying on CoJP (Section 5), and redistributed to current
nodes in case of rekeying (Section 6.2). As to the actual switch from
old to new permutation keys at the current nodes, a discussion on the
list with Malisa and Michael set the ground for a good approach to work
on. The last message on that topic is at [5].


At IETF 105, this work was put on hold, without proceeding to check for
WG adoption as requested by the authors. The reasons were: i) possible
termination of the WG later in 2019; ii) possible better belonging of
this work to IEEE rather than IETF.

Building on what mentioned at the mic at IETF 106 [6]:

1) I do believe that this document belongs to the IETF and its 6TiSCH
WG. The described solution intervenes precisely on the original
schedule, available from 6top or alternative protocols. Simply, the
permuted schedule is then used instead of the original one. Hence, the
proposed approach can be implemented above the MAC.

2) While IEEE 802.15.4 itself might be extended to similarly permute the
schedule, there is no chance that 802.15.4 will take care of
management/renewal of the necessary permutation keys. This would of
course limit and endanger the whole approach altogether. Instead, in
6TiSCH we can build on CoJP and the Parameter Update message from the
JRC used for link-layer traffic keys, to also provide and renew the
permutation keys (see above).

@Suresh: you asked about the document separating the schedule
permutation from the key management. This is in fact already the case,
as Section 4 describes the schedule permutation, while Section 5
describes the provisioning of the permutation keys to a pledge upon its
joining. The rekeying is now discussed in the small Section 6.2. If we
continue elaborating on the actual switch from old to new permutation
keys (see above), that point can definitely become an earlier functional
section of its own.


In summary, I am not in favor of closing 6TiSCH down yet, and I would
especially like to continue the work on this document, whose merits have
been recognized in the list, also by three positive reviews. At the same
time, as clarified above, I do believe that the IETF is its most
appropriate home, at the 6TiSCH WG.

Looking forward to your feedback and hopefully to progressing this work.

Thanks,
/Marco


[1]
https://gitlab.com/crimson84/draft-tiloca-6tisch-robust-scheduling/tree/master/test

[2] https://mailarchive.ietf.org/arch/msg/6tisch/AmdYIZJqwnLYkzXcyejWAfPRAGw

[3] https://mailarchive.ietf.org/arch/msg/6tisch/676O-xcWvHxRLjL-sSHQswS3mCk

[4] https://mailarchive.ietf.org/arch/msg/6tisch/a-SyZZWVluoTf-N8ky_D7Hu4lis

[5] https://mailarchive.ietf.org/arch/msg/6tisch/GdgNqNFGpcTRLCm4BI1O6BS36OE

[6] https://etherpad.ietf.org/p/notes-ietf-106-6tisch?useMonospaceFont=true

-- 
Marco Tiloca
Ph.D., Senior Researcher

RISE Research Institutes of Sweden
Division ICT
Isafjordsgatan 22 / Kistagången 16
SE-164 40 Kista (Sweden)

Phone: +46 (0)70 60 46 501
https://www.ri.se


Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch

Reply via email to