[6tisch] Comment to draft-wang-6tisch-6top-coapie-01.txt

[Tero Kivinen <[e-mail]>]

This draft plans to put the coap messages directly on the payload IEs
in the 802.15.4. The problem is that we only have 16 payload IE types
in total for 802.15.4, so reserving one number for coap might be hard.

On the other hand IEEE 802.15.9 needed to have solution to this same
problem, i.e how to multiplex upper layer data packet over 802.15.4
frames and how to fragment them so they can fit on the 802.15.4
frames.

The 802.15.9 is split in two layers, first one is the multiplexed data
service, which takes 16-bit MultiplexId and the data packet to be sent
to the other end, and it will fragment and deliver it to the other
end. The MultiplexId is used to separate different protocols using
this mechanism. One of those protocols is the key management protocols
providing KMP for 802.15.4, which is the second part of the 802.15.9.

Instead putting the coap messages in the payload IEs, it would
possible to allocate one MultiplexId for coap messages, and then coap
messages could be transmitted using multiplexing layer of 802.15.9.
This would also take care of fragmentting the messages (with PHY using
127 octet PSDU, the maximum size of the data packet is around 24kB).

The 802.15.9 recommended practice is already past the working group
letter ballots, and is starting its sponsor ballot soon. It is
expected to come out around the end of year. The draft will be
available on the ieee document store after the sponsor ballot starts,
i.e. very soon.
-- 
[e-mail]