AWR1843BOOST: Noise Floor Calculation in OOB Demo AWR1843

Thank you for the response Sir!

I would like to know how do I access the thermal noise then? 

The scenario is slightly different from a typical automotive one. I am detecting the presence and range of a concrete wall at a distance of say 10 m. Given the wide HPBW of the antenna in E and H plane, I am getting multiple targets in my range profile. I am trying to extrapolate the SNR to estimate the maximum range using the measurement results from a static test bench at a range of 10 m. I have no information on the normalized RCS of the wall. Given a grazing angle of 90 degrees from a distance of 10 m, I want to estimate the SNR so that I can use the hardware specifications of the radar and the radar range equation to estimate the normalized RCS. For that to culminate, I need the noise power/noise floor.

Please guide me through this.

Best Regards

Maaz Ali Awan

Hi unfortunately we don't have information about this use case.

Thermal noise would need to be measured in an evironment where there are no reflections. 

Thank you

Cesar

Greetings Sir!

Thank you for the information. I sought it better in the absence of an anechoic chamber to estimate the noise floor by orienting the radar FOV towards sky in an open field by mounting on a tripod stand.

Thank you for your guidance and support.

Best Regards

Maaz Ali Awan

Thank you

Yes, pointing the radar to the sky is a good method to measure the noise floor

Cesar

Thank you for the response.

However, if I am reading the ADC values while looking at the sky, how do I convert the raw I/Q values into noise floor in dBm? It is easy to compare the relative power from a target and power in noise window for CFAR in terms of dB. I am not sure how do I interpret the ADC values? 

Actually, I am interested in observing the practical noise floor as a function of noise BW(observation window) for a single chirp. Moreover, I want to extend to multiple chirps and how integration of chirps reduces the noise floor by averaging out as it follows the Gaussian Curve.

Please guide

Best Regards

Maaz Ali Awan

My understanding on calculating the noise floor:
The scenario is no targets with radar looking towards the sky.

I am trying to estimate noise power in the time domain.

Utilizing a 16-bit ADC resolution with a reference voltage of 1V pk-pk, the raw ADC values obtained in a radar system are converted from unsigned to signed integers. In the context of a complex baseband architecture, these signed values correspond to the equivalent voltages of the I and Q components. The instantaneous voltage for each I/Q pair is then calculated by taking the square root of the sum of the squares of the I and Q values. Subsequently, the Root Mean Square (Vrms) voltage is derived from these instantaneous voltages. With an impedance of 50 ohms, the average power (P) of the signal can be calculated, which is then converted into dBm to evaluate the noise floor effectively.

Please verify my understanding.

Best Regards

Greetings Mr. Cesar!

The noise floor estimation did not go as per the plan. The coupling signature sabotages the whole concept of estimating noise in time domain. I tried implementing a HPF in MATLAB to suppress the coupling signature but it did not work unfortunately.

Please guide

Best Regards

Turns out, the issue was only in the early chirps for the 1st antenna. However, the rest of the chirps were fine. However, I have estimated a noise threshold of --108.5 dBm when the theoretical noise should have been -121.9 dBm. This difference is because of the coupling signature. So now I have decided to reduce the Tx power to the lowest level...-7.5 dBm for a backoff setting of 20 dB