discuss-gnuradio
[Top][All Lists]
Advanced

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

Re: [Discuss-gnuradio] problem with ofdm bandwidth and sampling rate


From: Marcus Müller
Subject: Re: [Discuss-gnuradio] problem with ofdm bandwidth and sampling rate
Date: Wed, 03 Jun 2015 15:56:03 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.7.0

I'd also like to point out that you should use an oscilloscope to verify that a sine with a 17 kHz frequency has the same peak-to-peak voltage as one with the same digital amplitude at 7 kHz -- sound cards do have nice DACs (bitdepth-wise), but the analog filtering that you must do after DAC'ing tends to be implemented in a one-stage passive way (at least for the cheap consumer-grade sound cards I've encountered), which will not make your spectrum pretty.
However, you said there's an existing Matlab implementation -- was the "Soundcard channel" measured with respect to frequency selectivity?

Best regards,
Marcus

On 06/03/2015 03:12 PM, Wunsch, Felix (CEL) wrote:
Hi,

please remember to always reply to the list.

If you mix your baseband signal in a way that no part of your spectrum includes negative frequencies BEFORE doing the complex-to-real conversion, this should work. At least from a signal processing perspective. I never used my soundcard as a transmitter, so I can't really help you with that.

--Felix

On 03.06.2015 15:02, address@hidden wrote:
hi felix,
 
>>I implemented a transmitter for the DRM(30) standard in GNU Radio and published it on github (https://github.com/kit-cel/gr-drm).
 
i have already installed your transmitter and tested it with DREAM. it works perfect. excellent job!
 
if i have an expert right here i can be more specific for my project: the aim for this work is not to implement a DRM transmitter/receiver, it's to implement an OFDM transmitter/receiver. previous implementations from my colleagues are done in Matlab. now i have to implement this work with an open source software (GRC) because there were problems with licensing and changed or removed functions in higher versions of Matlab. they used two computers with their sound cards, one for tx and one for rx. that's the reason why a center frequency of 12khz is used. the signal, slightly specified by the DRM standard (RM A, SO 3 --> bandwidth 10khz, occupied carriers 228 (-114 to +114)) has the highest frequency at 17khz. this signal you can handle with your sound card.
 
to your comments:
 
1) i know there are only specific sample rates allowed for DRM but as written previously i will not implement the full standard. at the moment i use a random source block to get in some random data.
 
2) i don't have an USRP. i want to use a sound card to get my signal out to the world (at 12khz). is this possible by using a complex-to-float block and an audio sink?
 
thank you very much for your help and your answers!
 
 
alphonso
 
Gesendet: Mittwoch, 03. Juni 2015 um 04:57 Uhr
Von: "Wunsch, Felix (CEL)" <address@hidden>
An: "address@hidden" <address@hidden>, "address@hidden" <address@hidden>
Betreff: Re: [Discuss-gnuradio] problem with ofdm bandwidth and sampling rate
Hi Alphonso,

I implemented a transmitter for the DRM(30) standard in GNU Radio and published it on github (https://github.com/kit-cel/gr-drm).

To (partly) answer your questions:

1) DRM only accepts specific audio sample rates, make sure to read the standard documents carefully.
2) Mixing is usually done by the USRP. You as the user only deal with the complex baseband centered around DC. 12 kHz would also be a VERY low carrier frequency, short-wave radio usually uses frequencies between 3 and 30 MHz.

--Felix


 
On 03.06.2015 12:57, address@hidden wrote:
hello,
i'm trying to implement an ofdm transmitter (and afterwards a receiver) for demonstration. the parameters should match the drm standard. so my first try was to implement the transmitter with the following main parameters:
 
ofdm transmitter block
fft length = 256 (--> fft length >= occupied carriers)
cyclic prefix = 64 (--> 256/4)
occupied carriers = 228 (--> (-114, -113, ..., 113, 114))
 
for this try no pilot carriers and symbols and no sync words. everything works fine but there is a problem with the spectrum of this ofdm signal. from older posts i have the following equation to set the bandwidth (bw) of my ofdm output signal:
 
bw = sampling_rate * occupied_carriers / fft_length
 
with the parameters above and the wish of a bandwidth of 10khz (to match the drm standard) i will get the following sampling rate:
 
sampling_rate = bw * fft_length / occupied_carriers = 10khz * 256 / 228 = 11.228khz
 
now i have two questions:
 
1) if there is an input source with a sampling rate higher than 11.228khz (maybe from my sound card (44.1khz) or a wav file (48 khz)) should i use a resampler block because there is a problem with nyquist? how can i handle this problem?
 
2) the same problem goes with the output. if i mix up the ofdm signal to a carrier frequency (for example 12khz) how can i handle this? there is also a problem with nyquist.
 
thanks in advance!
 
alphonso
 
-- 
Karlsruhe Institute of Technology (KIT)
Communications Engineering Lab (CEL)

Felix Wunsch, M. Sc.
Research Associate

Kaiserstraße 12
Building 05.01
76131 Karlsruhe

Phone: +49 721 608-46276
Fax: +49 721 608-46071
E-Mail: address@hidden

www.cel.kit.edu

KIT -- University of the State of Baden-Württemberg and
National Laboratory of the Helmholtz Association


-- 
Karlsruhe Institute of Technology (KIT)
Communications Engineering Lab (CEL)

Felix Wunsch, M. Sc.
Research Associate

Kaiserstraße 12
Building 05.01
76131 Karlsruhe

Phone: +49 721 608-46276
Fax: +49 721 608-46071
E-Mail: address@hidden

www.cel.kit.edu

KIT -- University of the State of Baden-Württemberg and
National Laboratory of the Helmholtz Association



_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio


reply via email to

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