|
From: | Isen I-Chun Chao |
Subject: | Re: [Discuss-gnuradio] Successful transmission in gr-ieee802-11 with adding extra samples (before OFDM preamble) |
Date: | Sat, 29 Nov 2014 06:49:47 -0500 |
On 2014-11-29 12:34, Isen I-Chun Chao wrote:
Okay, I got it.
One more quick question.
I have been pretty confused about what the sync long and sync short are,
respectively, responsible for?
Sync short is for frame detection and searches for the cyclic pattern of the short preamble (autocorrelation). Once a (potential) frame is detected, sync long correlates sync_length samples with the long preamble and searches for peaks to align OFDM symbols.
Thanks.
/Best Regards,
Isen I-Chun Chao/
On Sat, Nov 29, 2014 at 6:29 AM, Bastian Bloessl <address@hidden
<mailto:address@hidden>> wrote:
On 2014-11-29 12:28, Isen I-Chun Chao wrote:
Hi Bastian,
Thanks for your reply.
The peaks you mentioned is like the correlation result?
Yes, exactly
/Best Regards,
Isen I-Chun Chao/
On Sat, Nov 29, 2014 at 6:21 AM, Bastian Bloessl
<address@hidden <mailto:address@hidden>http://www.ccs-labs.org/~__bloessl/ <http://www.ccs-labs.org/~bloessl/><mailto:address@hidden <mailto:address@hidden>>> wrote:
Hi Isen,
On 2014-11-29 07:01, Isen I-Chun Chao wrote:
Hi,
I am using gr-ieee-802-11 and putting a custom block,
which is
responsible for adding a 90-sample preamble at the head
of incoming
sample stream, right after the output of WiFi PHY Hier,
as attached
figure, generator.png. So the sample stream is changed
as attached
figure, adding_preamble.png.
However, in the case of use of transceiver.grc, I can still
successfully
decode received data, which is keep printing out "Hello
World!".
Does anyone know why adding extra samples at the head of
transmission
samples does not affect the receiving results? Does it
because
somewhere
in Rx side detect the OFDM preamble so the payload can
be normally
processed without being affecting by the extra samples
placed before
OFDM preamble?
The sync long block does matched filtering with the long
preamble
and searches for peaks in a configurable window (sync length
parameter). Looks like even if you add 90 symbols the peak
that the
block is looking for is still in the window.
Best,
Bastian
--
Dipl.-Inform. Bastian Bloessl
Distributed Embedded Systems
University of Paderborn, Germany
--
Dipl.-Inform. Bastian Bloessl
Distributed Embedded Systems
University of Paderborn, Germany
http://www.ccs-labs.org/~bloessl/
[Prev in Thread] | Current Thread | [Next in Thread] |