| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Hi Young Hoon, Thanks for the question and comments. Please see my response below in-line. Regards, Yunbo 发件人: Young Hoon Kwon [mailto:younghoon.kwon@xxxxxxx]
Hi Yunbo, Regarding the use of PIFS on a link that frame exchange is successful (SP1), we already have a similar discussion (487r5) and the SP result was:
SP#2: Do you support to define a mode that when doing simultaneous frame exchanges with STA MLD without STR capability, the inter-frame space between the ending time of the short responding PPDU and the starting time of the following soliciting
PPDU may be more than SIFS and no more than TBD time?
·
Note: it may be required to do ED sensing when the IFS is longer than SIFS according to EU regulation.
Y/N/A/No answer: 12/36/33/16 Even though we are still positively thinking about the use of PIFS for this case, it looks the group’s feedback on this case was quite not-supportive. [Yunbo] Yes, the SPs are similar. Maybe liwen mentioned too many scenarios (e.g. transmit to another MLD after wrong responding), we can further discuss with more people under specific
case. If someone gets better solution, we can further discuss. For me, I didn’t get other better solution for the cases I mentioned in slide 5 and 7 of 1062r0. And, for your SP2, I have one clarification question. Is the SP2 targeting the case that the AP MLD does not receive any acknowledgement on all links? Or, is it for the case that the AP MLD receives an acknowledgement
at least on one link? [Yunbo] It can work for both cases, regardless of whether acknowledgement is received or not on another link. If you see any problem, please let me know. BR, Young Hoon From: Liyunbo <liyunbo@xxxxxxxxxx>
Hi Young Hoon, Thanks for response. I think we don’t have conflict design until now. But we need to discuss how to harmonize the text to make the rule clean for different cases. Please see my response
below. Regards, Yunbo 发件人: Young Hoon Kwon [mailto:younghoon.kwon@xxxxxxx]
Hi Yunbo, Thanks for letting us know your proposal. It looks interesting and I’ll give my feedback soon. Regarding your comment on my SP, I think your intention is to extend the proposed rule to be applied to a case that the AP MLD even does not receive immediate response on another frame. (but at least PHY-RXSTART.indication
is received.) [Yunbo] you are correct. If we go back to the original proposal, one of rationale that I would like to relax current PIFS recovery rule is because an AP MLD has at least maintains a successful frame exchange on at least one link. (It is similar
to 11ax OFDMA acknowledgement case, where the TXOP owner considers the frame exchange successful if the TXOP owner receives successful response from at least one scheduled STA.)
[Yunbo] I think they are quite different. OFDMA acknowledgement is happens in single link. But what we discuss now is for two links, we should decouple the operate in two links as
much as possible if it doesn’t has strong reason to couple them. Now, you want to relax the current PIFS recovery rule even for the case that there’s no link that successful recovery is received.
I’m not totally against this idea, but at least I think it needs further discussion as different members may have different level of flexibility in relaxing current PIFS recovery rule for non-STR MLDs. [Yunbo] we can discuss more, and see opinions from more people. I still prefer the delayed PIFS covers both cases by now. Unless we can find a stronger reason to distinguish the
two cases. So, I would like to have separate discussion on your proposed modification, which can be done together with your other ideas in 1062r0. Good thing is that you are not against the behavior shown in the SP when the AP MLD fails on one link but receives response on another link successfully, right? [Yunbo] I am not against your design, but want to have a fully discuss before you run the SP. Because the delayed PIFS also related to other cases. Thanks, Young Hoon From: Liyunbo <liyunbo@xxxxxxxxxx>
Hi Young Hoon, Thanks again for waiting. I uploaded my presentation in the server, please have a look and let me know your comments. After technical discussion we can further think about whether
need to merge or modify our SPs. https://mentor.ieee.org/802.11/dcn/20/11-20-1062-00-00be-error-recovery-for-non-str-mld.pptx After review your SP again, I have one comments: Whether below text in yellow is not needed in your SP, or can extend to PHY-RXSTART.indication of response frame is received in another
link. Because when response is failed on another link, AP MLD should also allowed to continue its transmission on the this link. And AP may follows the same procedure whether response on another link is failed or successful. (When PPDUs are failed in both
link, the procedure can covered by PIFS recovery rules in current spec. my comment doesn’t cover this case)
•
Do you support the following transmission sequence for the constrained multi-link operation:
–
When an AP MLD obtains TXOPs on multiple links and transmits frames to a non STR non-AP MLD soliciting immediate response on the multiple links, and intends to align
the ending time of DL PPDUs during the obtained TXOPs:
•
If the AP MLD does not receive an immediate response successfully on a link that is not a first frame exchange within the obtained TXOP of the link,
and if the AP MLD receives an immediate response on at least one another link, the AP MLD can continue its transmission on the link within the obtained TXOP TBD (e.g., SIFS or PIFS) time after the
failed reception of the immediate response if the channel is idle.
•
CCA mechanism on the link is TBD. Regards, Yunbo 发件人: Young Hoon Kwon [mailto:younghoon.kwon@xxxxxxx]
Hi Yunbo, Sure, let me know whenever you are ready. Best, Young Hoon From: Liyunbo <liyunbo@xxxxxxxxxx>
Hi Young Hoon, Thanks a lot to defer the SP to have a further discussion. I have several cases that want to discuss with you and all other people. It is a little difficult to discuss only through
text in e-mail, could you give me a couple of days to prepare a presentation, and then we further discuss base on that? It may would be more efficient to discuss with figures and clear motivations. Regards, Yunbo 发件人: Young Hoon Kwon [mailto:younghoon.kwon@xxxxxxx]
Hi Yunbo and Ming, Based on our discussion during the TGbe MAC call, let me initiate another round of email discussion regarding 20/427 synchronous multi-link operation. To the best I remember, your comment was to figure out the overall solution considering all possible operation scenarios. Let me first explain what is covered by this SP.
-
When an AP MLD receives immediate response frame on one link only, and No RXSTART indication received in the other link.
-
When an AP MLD receives immediate response frame on one link only, and RXSTART indicated received but decoding failed on the other link. And, these operation are limited to the case that AP has ability of relatively fast cross-link communication. So, can you please let me know which operation scenario do you want to consider on top of the operations that current SP covers? Thanks, Young Hoon 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 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 |