Beamforming issue using ULA FMCW radar

Hello everyone,

I have created a MIMO FMCW radar with 3 transmit antennas and 4 receive antennas. I am currently simulating a radar scene in a free-space environment with 6 different targets, while all three transmit antennas operate simultaneously.

I would like to verify the usefulness of applying an analog beamformer (phase-shift beamforming) at the receiver. To do this, I compare the range-Doppler magnitude maps obtained from the 4 receive antennas without beamforming to those obtained when phase-shift beamforming is applied before amplification and dechirping.

However, regardless of the steering direction I choose — whether it corresponds to a target direction or not — the magnitude at each of the 4 receiver antennas remains higher in the case without beamforming than in the beamformed case.

Could someone help me understand why this happens, or point out what I might be missing in my implementation?

Thanks in advance for your help.

[George]

Hello - would you be able to share the code that is producing this unexpected result? I think it will be difficult to say what the issue is without some more information.

Answer 1

[Kais]

Hello George,

Thank you for your reply. I have actually found the source of the issue.

The behavior was related to the WeightsNormalization property of the phased.PhaseShiftBeamformer object. I was comparing results obtained with different normalization options ("Distortionless" versus "Preserve power").

With "Distortionless", the beamformer normalizes the weights to maintain unity gain in the steering direction, which reduces the overall output magnitude compared to the unbeamformed signals. When using "Preserve power", the output power is preserved and the results are more consistent with my expectations.

So the lower magnitudes I observed after beamforming were due to the weight normalization method rather than an error in the beamforming implementation itself.

Thanks again for your willingness to help.