This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AWR1443BOOST: Radar intensity measurements

Part Number: AWR1443BOOST

I am attempting to measure the intensity of returns from an object with the goal of predicting the intensity of a return from the same object at a larger distance. My experiments have given results that don't match theory.

I am recording data from an xWR1443 using ROS. Once I have recorded data using ROS, I am processing it by manually selecting which points are associated with my object and summing up the energy in each point. Because the energy reported by ROS is on a logscale, I convert it to a linear scale before summing up, then convert back. (For the energy of each point associated with the object, x, Sum = Sum+10^(x/10), TotalEnergy = 10*log(Sum)). Is this an appropriate way to determine the amount of RF energy being reflected from a target?

I have performed an experiment where I measure the intensity of the returns from a 6" trihedral corner reflector at various ranges and noticed that the TotalEnergy associated with the object does not decrease according to what I would expect based on the radar range equation Energy  = (Pt * Gt * Gr * λ^2 * σ^2) / ( 4 * π )^3 / R^4.  Based on this equation I would expect TotalEnergy to decrease at a rate of 1/R^4, but I am seeing a decrease that more closely matches 1/R^2.

What reasons could there be for the sensor behaving this way? I realize that some energy may not be represented because some returns are filtered out. My cfarCfg is set to 854 (10 dB), so I am hoping not much energy is being removed this way. Still I am unsure why the range dependence I am finding is different than what theory predicts.

Thank you,

Andrew

  • Andrew,

    This question will require some further investigation. Please a few days for a response from TI. We will need to confer with some of the experts on our systems teams.

    Regards,
    Kyle
  • Andrew,

    After a few discussions with our systems team, I have a few follow-up questions.

    What kind of object are you using for your measurements? Are you using an ideal reflector like a corner reflector? If not, your measurements could be affected due to the RCS of your target object.

    Regards,
    Kyle
  • I am using a 6" trihedral corner reflector with an RCS of 37.2m^2 (15.7 dBsm).

  • Data was sampled from a 6" trihedral corner reflector with an RCS of 37 m^2 (15.7 dBsm) at ranges varying from 5-30m.
  • I ran another test which I think will be easier to discuss. I recorded data for the same corner reflector (6" Trihedral CR) at a range of 4 meters using different chirp configurations. Each chirp configuration is identical except the RX gain is increased from 24dB to 48 dB in increments of 6dB. The resulting point cloud is recorded and aggregated. All points associated with the corner reflector are converted to a linear scale, summed, then converted to a logarithmic scale.

    I would expect the intensity of the returns to increase by the same amount as the RX gain, however I see that the gain increases by 3 dB for each 6 dB increase in RX gain. In the attached screenshot the five steps correspond to an RX gain of 24, 30, 36, 42, and 48 dB from left to right. Why would I be seeing this relationship?

    Thanks,

    Andrew

  • Andrew,

    This seems to be an issue with the processing and not the with antenna. Your CFAR threshold could be filtering out energy that would affect your graphs.

    Could you upload the chirp configurations that you are using for testing?

    Regards,
    Kyle
  • I have attached a copy of the radar configuration file used. For each reading the profileCfg command was modified.

    Measurements were taken for each of the following chirp profiles.

    profileCfg 0 77 17 7 80 0 0 50 1 208 2888 0 0 24
    profileCfg 0 77 17 7 80 0 0 50 1 208 2888 0 0 30
    profileCfg 0 77 17 7 80 0 0 50 1 208 2888 0 0 36
    profileCfg 0 77 17 7 80 0 0 50 1 208 2888 0 0 42
    profileCfg 0 77 17 7 80 0 0 50 1 208 2888 0 0 48

    GainTest.cfg