Seongho,
Thanks for the comments.
Some responses:
1) During NPCA operation, successful PPDU
transmissions/receptions are obviously a bonus time to utilize the momentarily
empty resources. Therefore, it is very unreasonable for non-AP STAs to set the
MU EDCA Timer due to UL TB PPDU transmission during NPCA operation while MU
EDCA Timer is 0
I disagree.
The MUEDCA timer exists for the purpose of allowing a non-AP STA to perform untriggered UL TX when the AP is not providing enough triggers for UL to that STA.
Whether the AP grants a trigger for UL to a non-AP STA on channel A or channel B or channel C is completely immaterial to the discussion.
The only thing that the STA cares about is "Am I getting enough triggers?"
Therefore, I believe that the use of the MUEDCA timer across subchannels is completely appropriate.
NPCA provides more effective BW available for the AP to meet its commitments to the STAs by giving it more opportunities to transmit triggers to them.
2) If an NPCA non-AP STA successfully transmits any
PPDU(s) including OM Control subfield with UL MU Disabled subfield set to 1 or
UL MU Data Disable subfield set to 1, there no way to prevent untriggered UL
transmission during NPCA operation for that STA. Am I correctly understanding
This seems to be correct.
MU EDCA has a philosophy of fairness for UL
transmission, while NPCA uses it to achieve operational efficiency such as
better coordination/management in the NPCA primary channel by blocking
untriggered UL transmission. Since they have different purposes, I think it is
more appropriate to separate the MAC variables related to MU EDCA Timer during
NPCA operation from BSS primary channel
I disagree with the statement that MU EDCA exists to promote fairness.
MU EDCA exists to allow the AP to determine access according to its understanding of QoS objectives, and the result is very likely, unfair.
I do, however, agree that the use of MU EDCA in the NPCA channel is different from its use in the BSS primary channel.
Whether it is an appropriate use of MU EDCA is a different question and a much larger audience would need to voice its collective opinion before any change can be made.
This is a significant technical issue that cannot be addressed through a two-way dialogue.
If you have an alternative proposal, feel free to create a presentation and have it queued for discussion in the TG.
3) redundant phrase at the end of the MU EDCA parameters section
I agree, the phrase will be deleted.
4) better to use NPCA STA and NPCA peer STA?
I do not think it is better, as I see nothing wrong with saying first STA and second STA.
I.e. philosophically, don't fix something that is not broken
And if you change to "NPCA peer STA" then everyone will want a definition for that.
5) As mentioned in CID 787, if an NPCA STA (i.e., a
first STA) is an NPCA AP, the NPCA AP shall not initiate a transmission during
the longest switching delay among the switching delays of all NPCA non-AP STAs
that are considered to be the recipients of the first ICF
Realized, based on this comment, that CID 787 did not have a resolution.
In response to the comment, I have added a "reject" resolution to CID 787.
The language already includes the possibility of mulitple recipients,so no change is needed.
6) what to do when an NPCA non-AP STA that is
operating NPCA receives an ICF from its associated AP with the NPCA Primary
channel Indication field not set (i.e., set to 0).
Agree.
New line added in the NPCA transmission section.
To unsubscribe from the STDS-802-11-TGBN list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-TGBN&A=1
