Thread Links | Date Links | ||||
---|---|---|---|---|---|
Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
Hi Jeongki, Thanks for sharing the summary. Let me share my opinion on your Q1. Base on your answer of Q2, the AP needs to first obtain an TXOP and send BSRP to collect the buffer status at non-AP STA side. When AP send a CF-END
after BSRP/BSR, AP is in an obtained TXOP. So it follows below rule in spec. “When a STA gains access to the channel
using EDCA and empties its transmission queue, it may transmit a CF-End frame provided that the remaining duration is long enough to transmit this frame.” Regards, Yunbo 发件人: Jeongki Kim [mailto:jeongki.kim.ieee@xxxxxxxxx]
Hello Alfred and all, Thank you very much for your comments and questions on this (CF-end for an overlapping quiet interval)
during the last call. I summarized our discussions as follows:
·
Q1: Does the current spec already cover that operation?
·
A1: I don’t think the spec clearly describes the operation.
According to the current spec, the TXOP holder can transmit CF-End in its own TXOP and AP may further transmit CF-End after receiving CF-End. The related spec text says
“When a STA gains access to the channel
using EDCA and empties its transmission queue, it may transmit a CF-End frame provided that the remaining duration is long enough to transmit this frame.”. I don't think that the NAV set by quiet interval
is its own TXOP by EDCA.
·
·
Q2: How does the AP know the non-AP STA’s buffer status?
·
A2: One option is AP can transmit BSRP to the STA. AP can know no buffered traffic of the STA based on the received BSR.
·
·
Q3: What if the STA enters the doze state for the quiet interval? In that case, the STA will not receive CF-End.
·
A3: In the baseline spec, I did not see any description that the STA receiving quiet element should enter the doze state for quiet intervals. The current
spec just says that the STA shall set its NAV based on the quiet interval. Although it’s possible by implementation, I don’t think all STAs
receiving quiet element will enter the doze state.
·
·
Q4: Why do we need this operation for overlapping quiet interval?
·
A4: As I mentioned in above A3, in the baseline spec, the STA receiving quiet element shall set its NAV based on the quiet interval. It’s
the purpose of this operation. In 11.8.3 of REVme, “The NAV shall be set by a non-VHT STA in the BSS for the length of the quiet interval established by a Quiet element.”
·
·
Q5: Do we need this operation for 1TU overlapping quiet interval?
·
A5: I agree that the overlapping quiet interval is smaller than the normal quiet interval. Basically, I think the spec should provide a mechanism that
11be AP can transmit a CF-END during the overlapping quiet interval for resource efficiency. Anyway, non-AP STAs will be able to have more transmission opportunities through this operation. I expect CF-end procedure will be less than 200us. Then, we have 800us
room. Note that whether(/when) AP transmits CF-End depends on AP’s implementation.
·
Q6: How about STA operation for this?
·
A6: The STA will follow the baseline rule in 10.23.2.10.
“In a non-DMG BSS, a non-S1G non-HE STA(11ax) shall interpret the reception of a CF-End frame as a NAV reset.” Please let me know if I’m
missing something. Any comments are welcome. Thanks, Jeongki To unsubscribe from the STDS-802-11-TGBE list, click the following link:
https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-TGBE&A=1
To unsubscribe from the STDS-802-11-TGBE list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-TGBE&A=1 |