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

Re: [STDS-802-11-TGBN] Regarding contribution 25/1445r1 on "NPCA optimizations"



Hi Klaus,


Thanks for clarifying.


We are not trying to claim that a very accurate calculation of gains using various features can be performed. However, an estimate depending on load and bandwidth conditions is possible and is important for the non-AP to make decisions while enabling various features.


On your 2nd point, the non-AP indeed has choices to make between various features when it has limited resources. Please also note that all these features except EMLSR are per-link. With the memory/hardware resources that a non-AP has, if it can support only one feature between NPCA/DSO on a certain link or if it has to choose the link on which it should enable these features, it will benefit from making an informed choice rather than a random one. This is relevant even in the special examples using NPCA and DSO which you cited although there are many other configurations possible considering different bandwidths, links, EMLSR and DPS. For example, even if the non-AP bandwidth is 80MHz within a 160MHz AP bandwidth, it is allowed to perform NPCA outside of its operating bandwidth. In some configurations (e.g. 80in320 DSO), the NPCA PC may not even align with the DSO subband leading to additional resource constraints. Also, as I mentioned, there can be multiple links on which these decisions need to be made while the resource constraints are specific to the device.


Regards,

Sindhu



On Wed, Sep 17, 2025 at 8:13 AM Klaus Doppler (Nokia) <klaus.doppler@xxxxxxxxx> wrote:
Hi Sindhu,

  1. Klaus Doppler: Would love to see some results

    1. I assume that this comment corresponded to the NPCA gain guidance topic as the other proposals have very  deterministic benefits.  Please let me know if that is correct. However, I think even this proposal has deterministic benefits in terms of whether the non-AP enables or disables a feature considering its resource constraints and expected gains? What is the aspect you would like to be evaluated? 


Klaus: Yes, I tried to understand what you are trying to do with the gain guidance. As we have seen in our simulations the achievable gains are not always that straightforward to quantify. OBSS load is one factor but not the only one to be considered. Secondly, I am not sure, I agree with the statement that non-AP STA have to decide between NPCA, DSO and EMLSR. For example, I see NPCA mostly beneficial when you have e.g. a 160MHz non-AP STA in a 160MHz BSS. DSO would not make much sense in that case since the non-AP STA already covers the whole BSS BW. If you have an 80MHz non-AP STA it cannot use NPCA anyway since the NPCA PCH is outside of its bandwidth. 

BR, Klaus


From: Sindhu Verma <000011381223f2e2-dmarc-request@xxxxxxxxxxxxxxxxx>
Date: Tuesday, September 16, 2025 at 3:23 PM
To: STDS-802-11-TGBN@xxxxxxxxxxxxxxxxx <STDS-802-11-TGBN@xxxxxxxxxxxxxxxxx>
Subject: [STDS-802-11-TGBN] Regarding contribution 25/1445r1 on "NPCA optimizations"

 
CAUTION: This is an external email. Please be very careful when clicking links or opening attachments. See the URL nok.it/ext for additional information.

 

Hi all,


I presented the document 25/1445r1 NPCA optimizations a few days ago. Initiating this email to answer some of the questions raised in chat and also to elicit more feedback.


In summary, there were 3 proposals in the contribution. Embedding comments and responses below each of them:

  1. BSS color exception for NPCA:

    1. Abhishek Patil “Regarding p2p, there is a 11ax rule that requires p2p to use the same BSS color as the infra AP. Please see 26.17.3.2 Initial BSS color. I also remember there is some text in another portion of 11ax that talks about p2p in general - i.e., not limited to TDLS. So the NPCA issue you (Sindhu) mentioned are unlikely to occur when STA has p2p.”

      1. This does not seem accurate. Section 26.17.3.2 speaks of aligning BSS colors only in case of TDLS links. There can be P2P configurations that are not TDLS

    2. Brian Hart “not sure why you'd risk having different BSS Colors on colocated cochannel APs”

      1. Are you suggesting all co-channel APs in a network should have the same color? What is the point of BSS color in that case? If they have the same color, they would qualify as the same BSS from the perspective of any device. There will be no NPCA in that case.

    3. Laurent Cariou “for your first issue Sindhu, isn't the backhaul link always trying to use the max BW, therefore not allowing NPCA to be used anyway” (Gaurang had a similar comment during the meeting)

      1. The bandwidth of the backhaul link cannot be guaranteed. It depends upon CCA, data volume besides and additionally, proprietary aspects of implementation. In the absence of any negotiation regarding which color(s) to exclude for NPCA, the fronthaul non-APs would not be able to distinguish between cases where the backhaul bandwidth is smaller (due to say CCA) and any OBSS transmissions with smaller bandwidth

      2. Even if the backhaul transmissions always used max bandwidth when primary is available, what about the case when the fronthaul bandwidth < max bandwidth and the backhaul link attempts NPCA? It is not precluded to perform NPCA over backhaul and there can be NPCA transmissions even when the intended recipient is busy on the fronthaul on a smaller bandwidth. This again could be due to various reasons like client configuration, data volume, CCA, etc

    4. Jay Yang “isn't the backhaul link operating on a seperated channel from the fronthual in practise?”

      1. Not necessarily and even more rarely, in case of multilink operation where one or more links can be common between the backhaul and fronthaul

    5. Binita Gupta, Liwen Chu and Liangxiao Xin: Please send across your questions/comments


  1. NPCA gain guidance:

    1. Klaus Doppler: Would love to see some results

      1. I assume that this comment corresponded to the NPCA gain guidance topic as the other proposals have very  deterministic benefits.  Please let me know if that is correct. However, I think even this proposal has deterministic benefits in terms of whether the non-AP enables or disables a feature considering its resource constraints and expected gains? What is the aspect you would like to be evaluated? 

  2. NPCA parameter selection assistance:

    1. Shravan Kalyankar: Protocol for feedback

      1. We can discuss the next level details of how to send feedback once there is some consensus on the addition of this feedback. In our offline discussions, you also mentioned your proposal on indicating failure rate on the current NPCA PC. We can discuss that further. However, it is not possible for the AP to deduce the causes of failure or the remedial changes in NPCA PC or NPCA disabled subchannels bitmap just based on this information. 



Please add any other comments/questions in case they are missed out above.


Regards,

Sindhu



To unsubscribe from the STDS-802-11-TGBN list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-TGBN&A=1


To unsubscribe from the STDS-802-11-TGBN list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-TGBN&A=1