thanks very much. There is one more question. Please see inclined.
On 02/07/2017 22:20, Tero Kivinen wrote:
Jiliang Wang writes:
> I also have several questions as listed below:
>
> 1. How to set the length (in time) of each cell. This should be related to the
> number of retransmissions, i.e., for a higher number of retransmissions, the
> cell length should accordingly be set longer? This may incur some efficiency
> issue?
No. Each cell contains exactly on transmitted frame, and the ack
related to it. If there is needs for retransmissions they will be done
on the later cell, i.e., the transmitter delays sending until there is
cell to the same recipient, and also if this is shared cell, it will
also use backoff alcorithm so it will not use the next shared cell,
but skip some of them etc. For if the next cell is dedicated link,
then it will use it for retransmission.
See section 6.2.5.3 TSCH CSMA-CA retransmission algorithm in
802.15.4-2015, especially figure 6-6 center part, which covers the
TSCH retransmissions which are not priority access.
There is step "Delay for random (2^BE - 1) shared links or until
dedicated link" there which will cause the delay.
The Timeslot length is set by sending EB with TSCH Timeslot IE having
different value for macTsTimeslotLength.
> 2. What is default channelslot offset of EB, on which a new node
> unjoined node should listen?
It is what is sent in the TSCH Slotframe and Link IE of the EB, and
the section 4 recommends that slotOffset is set to 0x0000, and
chOffset is set to 0x0000.
It is still required a initial channel to receive the very first EB? How to set this?
> 3. If neither K1 nor K2 is pre-provisioned, a joining node uses the secExempt
> mechanism (IEEE Std 802.15.4, Section 9.2.4) to process all
> correctly-formatted EBs.
> Here I found no secExempt mechanism in Section 9.2.4 in IEEE Std 802.15.4
IEEE Std 802.15.4-2015 section 9.2.4 step f) will check for secExempt.
Note, that this section has changed since 802.15.4-2011 and earlier
versions of the 802.15.4, so you need to get the 802.15.4-2015 for the
section numbers to match. The actual functionality was there in
earlier versions also, but there were some mistakes in the state
machine, and it was harder to read.
Note, that 802.15.4-2015 is available for free from the IEEE Get
program, i.e. go to the
http://standards.ieee.org/getieee802/download/802.15.4-2015.pdf to get
it.
> 4. “Many mechanisms exist for discrimination between networks, the details of
> which are out of scope.” According to current field setting, it is difficult
> to differentiate different networks. Does it mean EB contains other field that
> can be used for this purpose?
There is ongoing work to allocate IE for IETF, and that IE could be
used to provide for example the DAG root or something similar in the
EB, which would allow finding the proper network during the joining
process, especially if you have been part of the network before.
--
[email protected]
_______________________________________________ 6tisch mailing list [email protected] https://www.ietf.org/mailman/listinfo/6tisch
