sampling rate requirements in radar end-to-end simulations

I am slightly confused about the sampling rates needed for radar end-to-end simulations. In order to model physical effects of the environment on the radar signal. so If I am simulating a radar operating between, let's say, 2 to 4 GHz, I would be usng a sampling rate of 8GHz. right? If I do so, matlab runs out of memory trying to create the waveform I need. Specifically, I am using phased.SteppedFMWaveform object which requires specifying the sampling rate.

The other problem I have is, that phased.SteppedFMWaveform generates N stepped pulses when you specify N (number of steps). The problem is I need to start at a different frequency, let's say 2GHz. How could this be done?

Thanks,

Answer 1

[Honglei Chen]

For a bandpass signal, in theory the sampling rate only needs to be the same as the bandwidth.

Unfortunately right now the waveform can only starts at 0. Can you share some information regarding in what situation you want to start at a different frequency? As a workaround, you can modulate the output of the waveform with a complex sinusoid with offset frequency. HTH

[darkhorse]

well, I am trying to create a High-range resolution (HRR) profile for a remote sensing application. My current effort is to compare FMCW vs SFCW to achieve the same thing operating anywhere between 3-10GHz with a bandwidth of about 2GHz. so as I mentioned earlier, if the starting frequency is 3GHz, I would like to use a step delta of 20MHz and get to 5GHz with 100 steps within a pulse burst.

Even If I use a baseband signal, say from 0 to 2GHz, then sampling rate turns out to be 4GHz, which is already quite high and I hit memory errors. or should I only consider the individual pulse's bandwidth which is going to be less than 1MHz (considering the pulse period) when setting the sampling rate?

seems to me that I am missing something fundamental about the phased array system toolbox usage..

[Honglei Chen]

If you are talking about complex baseband, then I think 2 GHz should work although there is nothing wrong with 4 GHz sampling rate. The memory limit is a practical concern. Currently Phased Array System Toolbox supports only double so sampling at such a high rate does demand a lot of memory. You can try to free things along the way but there is probably no silver bullet to address it at this point.