| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Hi Wook Bong, Thank for your discussion. My main concern is that if members want to remove 80+80/160+160, the comparison between the non-contiguous PHY transmission and the non-STR MLO needs to be evaluated. Because for the non-STR MLO,
some info needs to be timely shared cross links in order to realize the PPDU alignment, the implement complexity is much higher than the one of non-contiguous PHY transmission.
Hence, a possible implementation manner is that a non-AP MLD only supports the STR MLO. When the self-interference between two links is severe, it will just disable one of them. In this case, the
non-contiguous PHY transmission may be used. Regards Guogang Huang 发件人: Wook Bong Lee [mailto:wookbong.lee@xxxxxxxxxxx]
Hi Rui, Guogang, Ross, Bin, and Younghoon, Thanks for discussion. I am kind of agree with Rui. First of all, it is not clear when 80+80/160+160 will be used on top of MLO. As far as I know, end PPDU alignment is already agreed in MLO which is aiming for non-STR MLD.
Also, there is ongoing discussion on start of PPDU.
If both agreed, then having extra 80+80/160+160 seems duplicated mode of operation. Second, as Rui mentioned, 80+80 is not used much (actually I didn’t see one yet). If this is not used, but defined, then non-AP STA may implement it without using it.
If a certain feature is not useful and used, then it will be better not to define in new generation WiFi. Several PDTs are depending on this decision. Obviously we don’t want to write some text for some mode, and then remove it soon. So, it is better to be decided whether we will support it or not.
Is it possible to run following SP one more time during the PHY call tomorrow? Do you agree that 11be does NOT define PPDU with non-contiguous signal bandwidth?
·
Non-contiguous signal bandwidth includes 80+80MHz,
·
This does not include punctured modes within 160
Best regards, Wook Bong Lee From: huangguogang [mailto:huangguogang1@xxxxxxxxxx]
Hi Rui, “The non-contiguous PHY modes have been defined in two generations, and chip vendors and market have made the choice not to use the modes. ” This
doesn’t mean that the non-contiguous PHY modes still will not be used in 11be. I agree that currently seldom STA products support the non-contiguous PHY. In my understanding, the main reason is to reduce the cost of STA side. But with the advent of multi-radio
non-AP MLD device, I think that the non-contiguous PHY mode will be used in some scenario.
In your contribution, you only consider the comparison with the STR MLO. Maybe the comparison with the non-STR MLO is also needed.
Regards Guogang Huang 发件人: Rui Cao [mailto:rui.cao_2@xxxxxxx]
Hi Guogang, Agree that 80+80/160+160 is already in the SFD. But the non-contiguous modes was added following 11ac/ax style without thorough justification.
We have some analysis in our contribution that the non-contiguous modes do not provide throughput benefits. “So it’s a better way to leave the choice right to chip vendors.“ The non-contiguous PHY modes have been defined in two generations, and chip vendors and market have made the choice not to use the modes. This is the main reason that we propose not to define non-contiguous PHY modes in 11be to avoid unnecessary complexity in spec.
MLO is a new approach to enable non-contiguous operation. 11be may define different schemes for chip vendors/market to choose. But for proven unused modes, we don’t need to carry them into future.
Thanks, Rui From: huangguogang <huangguogang1@xxxxxxxxxx>
Hi Young Hoon and Rui, Thanks for the comment. The baseline is that the 80+80/160+160 is already in the SFD. I just follow the current SFD.
For the non-contiguous PHY transmission (80+80/160+160), I still want to keep it. There are the following reasons;
•
In some regions, there exists only non-contiguous channels, rather than the contiguous channels.
•
Comparing 80+80/160+160 with non-STR MLO, the channel access is much simpler. Based on the off-line discussion, some companies don’t like the non-STR MLO. So it’s a better way to leave the choice right to chip
vendors.
•
Since some info needs to be timely shared cross links in order to realize the PPDU alignment, the implement complexity of non-STR MLO is much higher than the one of non-contiguous PHY transmission. Hence, from implementation point of view, I want to support either way, depending on the scenario. Regards Guogang Huang Huawei Technologies Co.,Ltd. 发件人: huangguogang
Hi Young Hoon, I know this situation. The main reason mentioned in DCN 1100 is that few 80+80 operating STA on the market. But I think if the device is the MLD device, this mode can be used.
I suggest to represent the contribution 1100r0 and run it again in the next
joint teleconference call. After that, I can run my SPs. How about it? Thanks, Guogang Huang 发件人: Young Hoon Kwon [mailto:younghoon.kwon@xxxxxxx]
Hi Guogang, Thanks for bringing up this issue to the table. Actually I also have a similar contribution (684r0) that was presented last time and my own SP is in line with your SPs. However, based on the off-line discussion, quite many members want me to hold off my own SP because it depends on PHY’s decision on allowed PPDU bandwidth. And, even though some SPs in PHY (e.g., SP1 of 1100r0) does not get enough support to not allow 80+80MHz and/or 160+160MHz PPDU, it looks there’s no clear decision made on this issue in PHY side
yet. Under this situation, I think we better not be in a hurry in deciding the Channel Width and CCFS field definition until the PHY discussion on this topic is cleared. How do you think? Thanks, Young Hoon From: huangguogang <huangguogang1@xxxxxxxxxx>
Hi all, I have upload a new version of DCN680. I prepare to run the following two straw polls in the next week. If you have any concern or suggestion, please let me know.
•
SP2. Do you support to define EHT Operation element with two CCFS subfields (EHT_CCFS0/EHT_CCFS1) to indicate channel configuration for EHT BSS?
•
SP 3. Do you support to use 3 bits of Channel Width field in EHT operation element to indicate the channel width for EHT BSS as following
–
0: 20
–
1: 40
–
2: 80
–
3: 160/80+80
–
4: 320/160+160
–
5~7: reserved Regards Guogang Huang Huawei Technologies Co.,Ltd. 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 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 |