Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

[STDS-802-11-TGBE] 答复: [STDS-802-11-TGBE] 11-22-0349-02-00be-cr-discussion-of-nstr-and-emlsr



Hi Minyoung,

 

Thanks for your questions, please see my response below in-line.

 

 

Regards,

Yunbo

 

发件人: Minyoung Park [mailto:mpark.ieee@xxxxxxxxx]
发送时间: 2022324 7:04
收件人: Liyunbo <liyunbo@xxxxxxxxxx>
抄送: STDS-802-11-TGBE@xxxxxxxxxxxxxxxxx
主题: Re: [STDS-802-11-TGBE] 11-22-0349-02-00be-cr-discussion-of-nstr-and-emlsr

 

Hello Yunbo,

 

As I commented during the call, you are not doing apple to apple comparison. A NSTR MLD has two fully capable radios (two independent PHY/MAC blocks) whereas a single radio non-AP MLD operating in EMLSR has one fully capable radio. The NSTR MLD is close to double the complexity/cost of the non-AP MLD that has the EMLSR capability. The complexity/cost has to be considered in the comparison.

[Yunbo] The performance is much easier to compare because it has clear measurement criteria, that is an important reason we show performance in the presentation. For complexity, it related to too much implementation issues, it is hard to compare. I think it may reasonable to say NSTR MLD close to double the cost of MLSR, but I don’t know how to compare with EMLSR. Because EMLSR has a full radio and a “half radio”, so it can receive initial control  simultaneously in two links. We don’t know what’s the complexity and cost of this “half radio”. Maybe you or some RF/PHY expert can give a detailed explain about how to implement this “half radio”, and compare it with NSTR MLD. For now, I don’t think it is reasonable to say that “The NSTR MLD is close to double the complexity/cost of the non-AP MLD that has the EMLSR capability”.

 

Another observation is that a NSTR non-AP MLD with two 1x1 radios is actually less spectral efficient than 2x2 non-AP MLD in EMLSR mode since the NSTR non-AP MLD is using two 80 MHz links with 1ss on each link whereas the non-AP MLD in EMLSR mode is using one 80 MHz link with 2ss. For a busy network environment with many OBSSs that are not synchronized on both links (i.e. busy/idle are not synchronized on both links), this becomes a bigger problem to the NSTR non-AP MLD since most of time the two links are not idle at the same time and only 1ss can be used on one idle link whereas for the non-AP MLD in EMLSR mode it can still use 2ss on one idle link.

[Yunbo] It is true for SU and large packet transmission. For small packet transmission. It is not a big difference between single or two streams transmission. If MU-MIMO is used, there will be no difference, because it can aggregate more STAs to reach the same total number spatial streams. Considering the initial control overhead, the performance will worse than NSTR.

 

I also couldn't understand clearly why delay results are so high. There will be many cases where a STA can sync to the medium by receiving a frame from its own BSS or OBSS, by transmitting an RTS based on the current 11be spec, perform CCA if a STA is not doing 2ss tx, soliciting uplink traffic with a trigger frame, etc.

[Yunbo] Base on my understanding, for low traffic load scenarios, the initial control and switch delay are the main reasons to introduce a large delay. For middle load scenario, the load of EMLSR will be larger than NSTR, because initial control before each transmission will occupy a significant portion of airtime, so it increase the load of EMLSR, and furtherly increase the collision rate during the contention.

 

 

Regards,

Minyoung  

 

On Wed, Mar 23, 2022 at 9:48 AM Liyunbo <00001846a2e5e0c1-dmarc-request@xxxxxxxxxxxxxxxxx> wrote:

Dear all,

 

Thanks for your discussion during the presentation. I initiate this email thread to further discussion for doc 22/349. Please let me know your questions and comments if you don’t have time to express them during the call.

 

Here are some response to Minyoung and Shubho’s questions.

 

Q1: Whether switch delay of EMLSR affect the blindness in UL short PPDU transmission?

A1: when EMLSR MLD do UL transmission on link1, whether link 2 will enter blindness mode depends on the total duration of switch delay period and UL PPDU length. Assuming the UL PPDU length is 50 us, a 22 us switching delay (22+50=72us, which is the duration of MediumSyncThreshold) will trigger link 2 enter blindness mode. So you can see that link 2 will enter blindness mode even it is a small switching delay and short UL PPDU.

 

Q2: For blindness, can EMLSR be same as NSTR?

A2: No. For DL data transmission, NSTR will not enter blindness mode because most BA will shorter than 72us. But EMLSR will enter blindness for sure, because the initial control frame exchange already larger than 72us. For UL short packet transmission, NSTR will enter blindness when UL PPDU length >=72us, while EMLSR will not enter blindness when UL length >= (72-swithing delay). Consider the PHY preamble and switching  delay, you can see that EMLSR will very easy to enter blindness even for short UL packet.

 

Q3: Why complexity is not considered in the presentation?

A3: The main purpose of this presentation is to show the performance of EMLSR and NSTR. Because performance is a very important factor for people to judge a feature. For complexity, I don’t think it is an easy to compare, it relates to a lot of implementation details. Here I want to point out one more benefit of NSTR. There are full radios on both links for NSTR, so it is very flexible for a non-AP MLD to operate as STR (when AP MLD set up two links with enough frequency gap) or NSTR (when AP MLD set up two links very close). But EMLSR MLD can not switch between STR and EMLSR even AP set up two links with enough frequency gap.

 

Regards,

Yunbo

 


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