| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Hi Pooya, Thanks for your following question. Please see my response below in-line (in
green). Regards, Yunbo 发件人: Pooya Monajemi [mailto:pmonajemy@xxxxxxxxx]
Thanks Yunbo for the reply. My reply below to your second point On Tuesday, June 9, 2020, 08:13:08 AM PDT, Liyunbo <liyunbo@xxxxxxxxxx> wrote:
Hi Pooya, Thanks for your question, please find my reply below in-line. Hope it can address your question. Regards, Yunbo 发件人:
Pooya Monajemi [mailto:pmonajemy@xxxxxxxxx]
Hi Yunbo, This straw poll passed in the telecon and there was not much chance to discuss it: "When AP MLD transmit a BU in one link to a non-AP MLD, if there is at least one additional buffered BU of any TID or management
frames that is mapped to this link by TID-to-link mapping or default mapping for the same non-AP MLD, the More Data subfield is set to 1, otherwise the More Data subfield is set to 0." - When the bit is set to 1, there is no indication about which link the BU may be transmitted on. Should the receiver MLD stay
awake in all links? [Yunbo] The More Data subfield is only applied to the link that the BU is transmitted, so it doesn’t need extra link indication.
Receiver MLD doesn’t need to stay awake in all links. This motion doesn’t have any impact on when to switch from doze state to awake state for a power save STA. A STA just follow current power save mechanism to awake on each link. This straw poll just gives
the rule that when AP MLD transmit a BU in a link to a power save STA that affiliated to a non-AP MLD in awake state, how to set More Date subfield to 1 or 0.
- When the bit is set to 0, there may be still BU's outside a common buffer and getting ready for transmission in another link.
In this case the receiver should not set the other link to doze based on the 0 value in the first link [Yunbo] When there is still buffered data getting ready for transmission in another link, and the TID of buffered data is mapped
to the transmitting link, the More Data subfield shall be set to 1 according to straw poll. So the issue you mentioned will not exist anymore. Actually, it is the main purpose or scenario the straw poll focus on. [Pooya] Here your assumption is that the AP's of the AP MLD have instantaneous knowledge of the status of the PPDU's in another link. The AP may set
the more data to 1 while the PPDU in the other link has already been transmitted and there is no more data. [Yunbo] The assumption is AP’s of the AP MLD has knowledge of the status of PPDU’s in another link, but it is not necessary to be instantaneous
knowledge. In the scenario you just mentioned, if AP set More Data subfield to 1 while there is no more date in other link. When there are more data on transmitting link, the AP can set More Data to 1 for the last PPDU on transmitting link, when there is no
more data on transmitting link, the AP can send a QoS Null frame with More Data to 0 to help the receiver STA to enter doze state.
I mentioned similar scenario in my presentation, an AP set the More data to 0 on a transmitting link because of there is still buffered data on
another link, but later the buffered data all transmitted on the other link (not schedule or retransmitted on the transmitting link), the AP also need to send a QoS Null frame with More Data to 0 to help the receiver STA to enter doze state. Thanks Pooya 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 |