discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Recover the signal


From: Marcus Müller
Subject: Re: [Discuss-gnuradio] Recover the signal
Date: Tue, 27 Oct 2015 17:53:38 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.7.0

Ah, so it's FFT of samples 0-127, then FFT of samples 128-255, then FFT of samples 256-383 and so on!
That has an interesting consequence: first of all, the (I)DFT of a Zadoff-Chu sequence is, in fact, also a ZC sequence. However, that doesn't mean that the DFT of a Zadoff-Chu sequence multiplied with a rectangular Window, is one, too:

let $s[n]$ be your Zadoff-Chu sequence, and let $l$ be your FFT length; judging from your figures, you don't actually use all $l$ carriers; let's say there are $m <l $ used carriers. The following Fourier transform equalities hold:

$s[n] \dft{l}
      S[k],\,n,k \in \{-\frac l2, \ldots, \frac l2 \}$, which also is a Zadoff-Chu sequence (i.e. magnitude $\equiv
      1$),
With your OFDM symbol being but the central $m$ out of $n$ carriers, I'd model the input of the IDFT as
$s[n] \,
      \mathrm{rect}_\frac m2[n] \dft{l}  S[k] * \mathrm{sinc}[k]$, which doesn't have Zadoff-Chu properties, so your signal in time domain will not have constant amplitude.
In theory, the inverse operation in the receiver should give you the same $s[n]
      \mathrm{rect}_\frac m2[n]$, maybe having a bit of noise, and maybe having a constant DC offset if any, IFF transmitter and receiver were synchronized.

What I assume is that the DC offset that you're seeing in the center peak is always the same, but the phase of the symbol start isn't; that leads to the DC offset always having the same phase, but the center symbol always a different one; if DC offset and center symbol happen to have the same phase, the peak is high, if they happen to oppose each other, you get cancellation.

Generally, depending on your bandwidth you might do something called offset tuning:
offset tuning
when you use "uhd.tune_request($f_\text{target}$,$f_\text{offset}$)" instead of directly using $f_\text{target}$ in the tuning frequency of a USRP block, you can use the USRPs capability to digitally tune by $f_\text{offset}$ in the DSP, moving the DC offset / LO leakage out of band.

Best regards,
Marcus


On 10/27/2015 12:15 PM, scott tiger wrote:
I received the signal in gnu radio sink, then take the file and divide it into 128 length sequences"because I used 128 IFFT in the transmitter". I made FFT to get the signal in frequency domain. Finally, I plotted the results.

Best regards
Maksim

On Tue, Oct 27, 2015 at 12:14 PM, scott tiger <address@hidden> wrote:
I received the signal in gnu radio sink, then take the file and divide it into 128 length sequences"because I used 128 IFFT in the transmitter". I made FFT to get the signal in frequency domain. Finally, I plotted the results.

Best regards
Maksim

On Tue, Oct 27, 2015 at 12:08 PM, Marcus Müller <address@hidden> wrote:
That's what I don't understand – how did you put them together?

I saved the received signal in a file, then I did further steps in matlab (FFT, gathering ...).
What **are** those steps?


On 10/27/2015 12:02 PM, scott tiger wrote:
Hi Marcus
of course in gnu, I used a band pass filter. But I have spikes in the center frequency of the signal.


That's what I don't understand – how did you put them together?

I saved the received signal in a file, then I did further steps in matlab (FFT, gathering ...).

Best regards
Maksim

On Tue, Oct 27, 2015 at 10:50 AM, Marcus Müller <address@hidden> wrote:
Hi Maksim,
In the receiver I made FFT and plot the correspond figure.
So, that's pretty clearly frequency domain of the receive signal, right?
So that might answer your question regarding DC offset: If there is DC offset, you'd see a constant spike at the center frequency. That's not really the case here, if I understand correctly.
The last figure which I sent it is the signal in the frequency domain which are repeated with each transmission "I put them all together to compare them."
That's what I don't understand – how did you put them together?

Best regards,
Marcus




On 10/27/2015 10:44 AM, scott tiger wrote:
Hi Marcus,
Y-Axis is the amplitude "abs(of the complex signal)".
X-Axis is not pure frequency domain or time domain, because the figure is drown from follow:
I generate ZC sequence "its amplitude equals to 1 in frequency domain" then I made IFFT and transmit the signal using USRP. The environment is a cable. I received the signal from another antenna of the same USRP. In the receiver I made FFT and plot the correspond figure. Since, I am the source file in the transmitter transmit the signal many times"repeat activated". The last figure which I sent it is the signal in the frequency domain which are repeated with each transmission "I put them all together to compare them."
I attached the same figure with more explanation "each black block is the signal in the frequency domain", but block 1 .....n is the same signal transmitted in different times.

Thank you for your reply
Best regards
Maksim


On Mon, Oct 26, 2015 at 4:47 PM, Marcus Müller <address@hidden> wrote:
Hi Maksim,

Could you keep this on the mailing list?

I don't fully understand:

> In fact, the figure shows repeated OFDM signal, each of it in frequency domain.

So you take the OFDM signal, and shift it in frequency domain, and then have N identical OFDM signals transmitted at the same time?
Can you clearly state what your X-Axis and what you Y-Axis are?


For example, I transmitted a zadoff-chu sequence which has a flat characteristic in frequency domain. The environment was a short cable with attenuation. The received signal also showed in frequency domain.
I attached it also "the figure shows the repeated sequences 2Mhz bandwidth in frequency domain". What I am curious about are spikes which appear usually in the center frequency? I thought may it is related some how with dc offset in USRP.

I don't understand this graph:
Maksim

What is the X-Axis, what is the Y-Axis?

Maybe you meant that you take values from a Zadoff-Chu sequence, IFFT them, thus generating an OFDM signal (which, by the way, is also a ZC sequence), add guard intervals and transmit them?



I attached it also "the figure shows the repeated sequences 2Mhz bandwidth in frequency domain". What I am curious about are spikes which appear usually in the center frequency? I thought may it is related some how with dc offset in USRP.
I'm really getting intrigued by what you observe :) but we'll really have to understand the graphs, which at this point, I'm afraid, I don't.

Best regards,
Marcus









reply via email to

[Prev in Thread] Current Thread [Next in Thread]