Thread Links | Date Links | ||||
---|---|---|---|---|---|
Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
Hi, Carlos,
Thank you for your further questions. Please see my reply below.
Regards!
Xiaoming.
From: "Cordeiro, Carlos" <carlos.cordeiro@xxxxxxxxx>
Date: Thu, 1 Aug 2013 22:53:21 +0000 To: Peng Xiaoming <pengxm@xxxxxxxxxxxxxxxxx>, "STDS-802-11-TGAJ@xxxxxxxxxxxxxxxxx" <STDS-802-11-TGAJ@xxxxxxxxxxxxxxxxx> Subject: RE: [STDS-802-11-TGAJ] Questions on 11-13/0433r1 Hello Xiaoming, Many thanks for your response. Please see a few other questions below. Thanks, Carlos. From: *** 802.11 TGaj - China Mill-meter Wave *** [mailto:STDS-802-11-TGAJ@xxxxxxxx]
On Behalf Of Peng Xiaoming (I2R) Hi, Carlos, Sorry for my late response. Please find my answer to your questions in the below email. I hope the replies can answer your questions. I look forward to hearing from you. Regards! Xiaoming. From: Peng Xiaoming <pengxm@xxxxxxxxxxxxxxxxx> Hi Carlos Thank you for your following questions although we did not have chance to discuss it face to face last week. I am on personal leave this week, I will try to get back to you on your questions ASAP next week. Sorry for my slow response. Regards Xiaoming
[Xiaoming]: In principal, the channel split is similar to the operation of channel switch. There is no need to redo association, RSNA, etc when a channel split takes place. [Cordeiro, Carlos] The fourth bullet in slide 11 states: “Non-AP/non-PCP STAs continue their packet transmissions in small band network in Channel 5 after tuning in and receiving the DMG Beacon frames sent out by PCP/AP 1 in its BTI, coupled with the necessary procedures like beamforming, associations and new schedules for service periods (SPs) and contention-based periods (CBAPs).” This is the reason for my question above. Perhaps this means association for new devices? [Xiaoming] Yes, the association means for new devices.
[Xiaoming]: Yes, this has been considered. Based on my understanding, these are two different methods.
[Cordeiro, Carlos] I believe 11aj should adopt both approaches. The more dynamic approach is similar to what is done in 11n/ac and would facilitate better coexistence. Do you agree? The method you mentioned here can only offer 2 logical channels over Chinese 59-64GHz bands. [Cordeiro, Carlos] I am not sure I understand what you are trying to say here. In both methods, the PCP/AP would anyway have to periodically beacon in the large channel. The difference is the duration of time
the devices remain on a small channel, but at the end of the day I don’t see a difference in terms of utilizing the small and large channels. Can you please elaborate? The proposed method in 11-13/0433r1 can offer at least three independent logical channels over Chinese 59-64GHz bands. To support at least 3 logical channel has been specified in the functional
requirement of 802.11aj. [Cordeiro, Carlos] Both methods can meet this requirement at the end of the day. Having the more dynamic approach would make things a lot easier implementation-wise, since there would be no need to change the BTI, A-BFT and ATI structure. [Xiaoming]: Based on your description here, I have limited understanding and do not know if it can meet the requirement. May I suggest if you can give a bit more details of the method
in a presentation slide format and schedule a time slot to present it to 802.11aj group? For example, you may consider to present it in the next conference call or the coming Sept meeting in Nanjing if you will attend? In here, I highlight some of my understanding
and questions. 1) The method only allow one PCP/AP to formulate a BSS. 2) all non-PCP/non AP STAs need to switch between large band channel and small band channel. 3) I am not sure which channel the PCP/AP will stay during DTI period? The large band channel
or small band channel?
Regarding the complexity of the proposed method, only PCP/AP needs to handle NPs. Non-PCP/non-AP STAs in small band channel does not need to know NPs. From overall network perspective, I agree with that there
is an increased overhead due to the additional BHI. However, I think this increment of the overhead is minimal.
[Cordeiro, Carlos]
1)
Please keep in mind that the PCP/AP can be a device as simple as a phone. Therefore, it does carry complexity. 2) It is not about the overhead, but about the implementation complexity. This reminds me of the WiMedia MAC, which uses a beacon period where devices send beacons back to back. Experience tells us that it was quite complex to make that work. [Xiaoming]: I have to say that this is quite different from WiMedia MAC, in which every device needs to send beacon and listen other's beacon with long period. However, in the method descried in 11-13/0433r1,
only two PCP/APs need to send the beacons and listen with each other. The complexity will be much lower. The good thing is that all non-PCP/non-AP STAs no need to switch between large band channel and small band channel.
[Xiaoming]: PCP/AP 1 will inform non-PCP/non-AP STA that was not the one issuing the Channel Split Request frame to switch to a small band (i.e., channel 5, for example as shown in slide 11)
through a Channel Switch Announcement element contained in DMG Beacon/Announce frames.
[Xiaoming]: In principal, for the channel switch from a large band channel to a small band channel, it will follow the similar way of the channel switch as described in 802.11ad specification.
That is to say, there is only one set of BF weight for a pari of STAs to hold. [Cordeiro, Carlos] That was not exactly my question. The question is: do you now if the weights for the large channel can be reused for the small channel? This would avoid re-beamforming. I am just wondering if the group has looked into that. [Xiaoming]: This is a good point. So far, the group has not looked into this. I should try to find an expert from the group to look into it. :-) The bottom line is we can reuse the
beaforming info until it requires the re-beamforming. Can you give me some good suggestion?
[Reply]: In slide 11, the length of Virtual BI (VBI) is the same as the length of BI on a large band channel. For illustration purpose in slide 11, it is shown that the length of VBI is the multiple
integers of the length of SBBI. However, the length of VBI can also be equal to the length of a SBBI. In your example, the typical length of a BI on a large band channel is 100ms, the typical length of a SBBI can also be 100ms. [Cordeiro, Carlos] I don’t quite understand how they can be exactly the same, since a S [Xiaoming] It looks the sentence is incomplete. Can you please complete your sentence? Thanks.
_______________________________________________________________________________
If you wish to be removed from this reflector, do not send your request to this reflector - it will have no effect.
Instead, go to
http://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-tgaj and then press the LEAVE button.
Further information can be found at:
http://www.ieee802.org/11/Email_Subscribe.html _______________________________________________________________________________
_______________________________________________________________________________
If you wish to be removed from this reflector, do not send your request to this reflector - it will have no effect.
Instead, go to
http://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-tgaj and then press the LEAVE button.
Further information can be found at:
http://www.ieee802.org/11/Email_Subscribe.html _______________________________________________________________________________
If you wish to be removed from this reflector, do not send your request to this reflector - it will have no effect. Instead, go to http://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-tgaj and then press the LEAVE button. Further information can be found at: http://www.ieee802.org/11/Email_Subscribe.html _______________________________________________________________________________ |