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

Re: [STDS-802-11-TGBD] [EXT] Re: [STDS-802-11-TGBD] About midamble design

HI Miguel,

 

Thank you for the clarification of your simulation settings.

 

Agree that randomizing pilot tones together with data tones should be relatively simpler. For single stream case, the potential channel interpolation and smoothing is more transparent to receiver.

 

Thanks,

Rui

 

From: ** STDS-802-11-TGbd -- Enhancements for Next Generation V2X.Task Group ** <STDS-802-11-TGBD@xxxxxxxxxxxxxxxxx> On Behalf Of Sebastian Schiessl
Sent: Friday, July 10, 2020 2:15 PM
To: STDS-802-11-TGBD@xxxxxxxxxxxxxxxxx
Subject: [EXT] Re: [STDS-802-11-TGBD] About midamble design

 

Caution: EXT Email

Hi Miguel,

 

I believe there should be no problems whether we randomize the pilots or not. The pilot sequence is already randomized, and we are not in danger of choosing a random sequence that reverts this previous randomization. I just think that the pilot value needs to be very clearly defined. 

 

In addition, some people have favored the option to rotate the data symbols as well in order to make the midamble randomization transparent to the receiver. My first intuition is that _not_ rotating the pilots will make this task easier for such an agnostic implementation, but this should be further investigated. This is a question to be answered by the designers who want to have such an agnostic implementation, I do not have an opinion on this personally. So I'm sorry for holding up your contribution, I'm very much in favor! I just wanted to know the details.

 

Best Wishes,

Sebastian

 

On Fri, 10 Jul 2020, 21:45 Miguel Lopez M, <00000f89ed8a0620-dmarc-request@xxxxxxxxxxxxxxxxx> wrote:

Dear Rui C, Rui Y and Sebastian S,

 

I checked my simulations and can confirm that I implemented randomization simply by multiplying the time domain waveform by +1 or -1. This means that the pilots where also randomized.  

 

After thinking a bit, I see no problem in randomizing the pilots (I think I had reached the same conclusion a few months back but my memory is not that good any more). One may leave the pilots unchanged, but it is simpler to randomize them as well.

 

My proposal is to leave the straw polls in 0862r1 as they are, perhaps clarifying that if randomization is implemented in the time domain then the waveform corresponding to an OFDM symbol (including the GI) is multiplied by +1 or -1, or something of the sort.

 

I’ll be happy to discuss any issues and to incorporate any changes or suggestions for improvement.

 

Best regards,

 

/Miguel

 

STDS-802-11-TGBD@xxxxxxxxxxxxxxxxx

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

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

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