• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 485 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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: angular blur and accuracy

Wembanyama
Intellectual 615 points
Part Number: AWR1843BOOST
Other Parts Discussed in Thread: AWR1843

hi,

when I use AWR1843boost to collect the raw data of one object, so that I can estimate the object's  Azimuth Angle.

I set one tx antenna enable and four rx antennas enable. and I know the angular resolution formular is:  lambda/(Mdcos(theta)), here, M= 4 for my settings in AWR1843.

based the formular when M=4, when the actual angle is 0 degree, then angular resolution is about 28.6 degree.

when the actual angle is 30 degree, the angular resolution is about 33 degree.

my question is:

according to the angular resolution formular, I found another formular about angular accuracy: 

so I want to know using awr1843 to detect object and estimate the angle, what value is the angular accuracy generally? give me some cases in number, thank you.

Regards,

Wembanyama

  • 0
    TI__Mastermind 20605 points

    Hi Wembanyama,

    I'm not sure if there is an equation that describes the error of angle detected as you have mentioned. I can check regarding this and whether we have characterized angular accuracy internally. In the meanwhile, can you tell me if you are looking at angular accuracy with respect phase error on the device, the EVM or the angle obtained in the OOB demo for the AWR1843?

    Regards,

    Kaushik

  • 0 in reply to Kaushik Gowda
    Intellectual 615 points

    hi,

    Kaushik Gowda said:
    can you tell me if you are looking at angular accuracy with respect phase error on the device,

     I am looking at angular accuracy with AWR1843boost by setting 1 TX enable and 4RX enable to detect one object near 10m distance. my task is to estimate one object's azimuth. I have a problem in processing my raw data to estimate object's angle. I found the angular error big. 

    Kaushik Gowda said:
    in the OOB demo for the AWR1843

    and now, I haven't used the demo yet.  actually I developed my current project( aiming to detect object, and obtain its' range and angle) on the TI's USRR demo.  I developed the project in CCS tool and setting one chirp in one Frame.

    by the way , based on your experience, do you think my current can achieve angle measurement?

    Regards,

    Wembanyama

  • 0 in reply to Wembanyama
    TI__Mastermind 20605 points

    Hi Wembanyama,

    Thank you for the details. Please find the following:

    1. Regarding the error you are facing with angular error, may I ask if you have performed phase calibration across the receiver channels?
    2. Yes. You should be able to quite easily detect objects at 10m with a 1Tx 4Rx antenna config on an 1843 device. You also mentioned that your project is based on USRR demo. How different is the detection performance compared to the USRR demo with whatever changes you have added to the same?

    Regards,

    Kaushik

  • 0 in reply to Kaushik Gowda
    Intellectual 615 points

    hi,

    Kaushik Gowda said:
    may I ask if you have performed phase calibration across the receiver channels

    unfortunately, I only processed the raw data to calculate range and estimate angle, I didn't do any phase calibration because I don't know how to take measures to perform phase calibration. could you give me more details, please?

    Kaushik Gowda said:
    You also mentioned that your project is based on USRR demo

    another question is : I want to know for my test detecting an object at 10m, range is ok, but here is  a critical point : I want to know if it can estimate an angle with a high accuracy?what level can be achieved?  and I know,  for my test , 1Tx 4 Rx ,the angular resolution is lambda/(M*d*cos(theta)), that's means M=4, the angular resolution is 28.6 degree, I'm right?

    for this poor angular resolution, how can I take measures to estimate angle? music algorithm?  

    Regards,

    Wembanyama

  • 0 in reply to Wembanyama
    TI__Mastermind 20605 points

    Hi Wembanyama,

    1. To perform calibration i.e obtain the phase mismatch coefficients, you can follow the instructions in the mmwave_sdk_userguide.pdf using the OOB demo.
    2. Yes, this would be the angle resolution. I would suggest you once go through the AoA estimation in the demo to understand how we compute the exact angle using angle FFT and the phase change computation.

    Regards,

    Kaushik