|
From: | Roee Bar |
Subject: | Re: [Discuss-gnuradio] USRP Complex Demodulation |
Date: | Mon, 16 Nov 2015 14:47:26 -0800 |
Thanks James, What do you mean by “USRP will generate the in-phase and quadrature components”? Are you referring to the USRP receiver or transmitter? I am feeding the USRP transmitter with a real only signal s(t)=cos(w0*t), why does the received signal has an imaginary component (i.e. Q(t) is not zero)? Thanks again, Roee On 16 Nov 2015, at 14:34, James Humphries <address@hidden> wrote: Hi Roee, The USRP is not expecting any particular type of data. The USRP is a direct conversion transceiver that operates on I and Q samples, so in theory, you should be able to generate/receive any type of signal. If you input a cosine to the USRP, in will generate the in-phase and quadrature components of that signal. For a simple cosine input, you get I(t)=cos(w0t) and Q(t)=sin(w0t) components. Yes, the USRP performs low-pass filtering, but this is to remove out-of-band noise and other signals. It is not to suppress the side-bands like you would see in a system that only samples the in-phase signal components. Remember, we are down-converting to baseband, so the low-pass filter actually filters from -BW/2 to +BW/2 (where BW is your total bandwidth, in this case, it is usually your sample rate). Those 'mirror' images don't exist in this system since both the positive and negative frequency components can be uniquely determined with the I/Q components of the signal. If you were to just zero out either the I or Q component, then you would see your mirror images in frequency again. -Trip On Mon, Nov 16, 2015 at 2:52 AM, Roee Bar <address@hidden> wrote:
|
[Prev in Thread] | Current Thread | [Next in Thread] |