• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 31 subscribers
  • Views 503 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

AWR1642: DOA algorithm different between SDK 3.x and 2.X

chianan chang
Intellectual 810 points
Part Number: AWR1642

Hi 

Does SDK 3.x and SDK 2.x using different algorithm? Because I use the same AWR1642 radar but different SDK demo application with the same configuration and test in indoor environment ( complicate environment with many static objects ), seems SDK3.x can get more accurate objects DOA and can prevent the clutter interference. 

  • 0
    TI__Guru 69660 points

    Hi,

    In SDK 3.x we have improved version of software/algo, so you may see a difference in the output.

    Regards,

    Jitendra

  • 0 in reply to Jitendra Gupta
    Intellectual 810 points

    Jitendra

    It is glade to hear SDK 3.x has improve of algo on DOA. But I have some questions:

    (1) I use a 8 cm range resolution ( parts of the configure like the attached photo, and we use 256 angle bins instead of original 32 s so the angle has about 0.44 degrees quantization value) to measure a big static carton, the second attached photo show an one shot of the detected points. You can see colume R the value =11.37m has 4 detected points and their azimuth angle are -1.79,-2.24,-1.34 and -1.79. According angle resolution formula with 8 virtual antenna, the angle resolution is 2/8 ~=14.3 degree. So there are has 4 detected point in the same R and V within angle resolution. Is that correct?

    (2) If TI's SDK 3.x can has the ability to detect multiple points withing the same Range , Velocity cell and within the angle resolution, can you let me know : (a)What kind of Algo is useed in this version (b) What kind of Algo use in SDK 2.x (c) Is there any limitation for example how many points can be detected within angle resolution or something else (d) Is there any parameter can set for this algo that I can tune it for different application (e) Where I can find the algo source code in the SDK 3.x?

    Thanks

  • 0 in reply to Jitendra Gupta
    Intellectual 810 points

    Hi Jitendra

    I ask new question 2 weeks ago but did not get your response yet.

  • 0 in reply to chianan chang
    TI__Guru*** 143834 points

    Hi,

    Sorry for the delay

    Will get back to you before the end of the week

    Cesar

  • 0 in reply to Cesar
    TI__Guru*** 143834 points

    Hi,

    Question (1)

    This should not be correct. In the same range bin only a point should be detected.

    Are you sure these are not reflections from the same object for different frames in time?

    thank you

    Cesar

  • 0 in reply to Cesar
    Intellectual 810 points

    Hi

       These points are in the same TLV detected objects.

        Do you have answer for my question (2)?

    Thanks.

  • 0 in reply to chianan chang
    TI__Guru*** 143834 points

    Hi

    Sorry for late delay

    (2) If TI's SDK 3.x can has the ability to detect multiple points withing the same Range , Velocity cell and within the angle resolution, can you let me know :

    [Cesar] We can detect two points in the same Range and Velocity bins. Detecting in the same angle resolution is not possible

    (a)What kind of Algo is useed in this version (b) What kind of Algo use in SDK 2.x (c) Is there any limitation for example how many points can be detected within angle resolution or something else (d) Is there any parameter can set for this algo that I can tune it for different application (e) Where I can find the algo source code in the SDK 3.x?

    [Cesar] Please see links to documentation

    file:///C:/ti/mmwave_sdk_03_02_00_04/docs/mmwave_sdk_module_documentation.html

    Thank you
    Cesar