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.

AWR1843BOOST: Threshold calculation method

Part Number: AWR1843BOOST

I checked the SDK user guide, but I didn't quite understand the definition of Threshold.

I know how to find the following, but is my understanding correct?

CUT (value to determine if it is a detected object) > (10^(Threshold/10)) + (noiseWin/(2^divshift))

In addition, which of the Range-Profile should I refer to for the CUT value?

  • Hi Hiroyuki-san,

    CUT refers to the Cell Under Test which is essentially every sample across your FFT data. The CUT is compared to the relative power of its surrounding cells to classify it as a peak that corresponds to an object. As the threshold against which the CUT is compared to is relative to its surrounding cells, it will have to be calculated iteratively for each CUT as the CFAR window slides across your samples.

    Regards,

    Kaushik

  • Thank you for your reply.

    From what you've said, is object detection measured by comparing the Relative Power values in each Range?

    Also, what is the relative power that will be the detection standard for that object?
    It would be helpful if you could point me to the document mentioned.

  • Hi Hiroyuki-san,

    Your inference in the first question is correct. About your second question, I'm not able to get a clear understanding here. Is it possible for you to rephrase the same or explain it in detail? What do you mean by "detection standard for that object"?

    Thanks & Regards,

    Kaushik

  • Thank you for your reply.

    I agree with the initial question.
    (Please let me know if there is a document that describes this idea)

    What I wanted to ask in my second question was to know which relative power value was used as the standard for comparison.

  • Hi Hiroyuki-san,

    We do not have a dedicated document explaining the CFAR algorithm. However, CFAR is a widely used algorithm and you should be able to find a lot of info regarding the same on the web across multiple sources. There are also many threads discussing different nuances concerning the same on this forum which can act as a learning resource and may answer some of the questions that you might come across.

    To your second question, there is no standard value for comparison. The threshold is determined experimentally and the value to compare against your CUT is found adaptively based on your CFAR window.

    Regards,

    Kaushik