phased array toolbox

I have started using Phased Array Toolbox. Works well enough. It would be nice if there were a blog for it.

I've identified 2 bugs so far...at least I think they are bugs.

1. The gain for a rectangular array seems to be off by 5dB 2. For beamforming, if you add noise, and then sum, it appears to give incorrect answer. - Matt

Answer 1

[Honglei Chen]

Hi Matthew,

Could you be more specific about the issues you encounter? Thanks.

[Matthew]

for the rectangular array...consider a square array with each side of length L. The textbook provides the maximum gain as G=4*pi*Area/(lambda^2). However the matlab calculation using phased.ArrayGain gives the maximum gain as a few dB less. Appears to be an error in the gain calculation.

I could provide the specific code if that will help.

[Honglei Chen]

Hi Matthew,

I think the name may be a little misleading but the array gain in Phased Array System Toolbox refers to the signal to noise ratio gain, i.e., the SNR improvement achieved by an array compared to a single channel. It is not quite the same as the antenna gain/directivity. Maybe a better name is array processing gain. If you still think it's off, I'd love to look at the code.

What about the beamformer case? Thanks.

[Matthew]

Hi Honglei,

After reading your answer, I examined the function radareqsnr. In this function, the radar equation appears to be straight out of Skolnik, and it takes the gain as an input parameter. So are you saying that we should not use the gain calculated by PhasedArray.Gain as an input into the function radareqsnr? That seems a bit odd.

Additionally, the result from beamforming seemed consistent with the result from radareqsnr.

So if you are using non-standard definition, you seem to be following it with standard definitions, which would be inconsistent.

Thanks Matt

Answer 2

[Matthew]

Beamforming- The problem I'm having with beamforming...is that if you perform beamforming and then add noise, it should end up with the same result as if you perform the operations in the reverse order. However it does not end up the same.

Answer 3

[Matthew]

Good news- for a single isotropic element, I am getting agreement between my calculation and the Matlab calculation !

Answer 4

[Matthew]

I am getting a disagreement with the radar calculation for an array. Consider a target of RCS 5m^2, an array of 30x30 elements at frequency 1.3GHz (spaced at l/2), peak power=20kW, range=60kW, noise temp=500K. The radar equation gives me a SNR=33.8dB. The Matlab calculation is giving me SNR=24.8 dB

Kind of an odd difference of 9dB.