• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 30 subscribers
  • Views 393 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.

IWR6843ISK: Long Range People Detection

Mike Schell
Prodigy 170 points
Part Number: IWR6843ISK
Other Parts Discussed in Thread: IWR6843

Can the beam forming/steering be modified to cover a +/- 60-degree area? Maybe six beams at 20-degree intervals?

I understand the current demo may be limited to 4 beams, but I assume this is just a software limitation. Either way, can the IWR6843 support beam steering out to +/-60 degrees?

Thanks-

Mike Schell

  • 0
    TI__Mastermind 37805 points

    Hi Mike,

    Give me a day to run the calculation to see if it would make sense. My fear is that the  In the meantime, can you explain a little bit about what you're trying to do by beamsteering? It would likely increase your range but limit your field of view for each individual angle (though scanning at many angles would compensate for this).

    Additionally, you may still get detections in a +/- 60 degree area with the existing demo, you may just have reduced range at the edges of the FoV. Have you tried using the demo as it exists today to see if it meets your needs?

    Best,

    Nate

  • 0 in reply to Nathan Block
    Prodigy 170 points

    Hi Nathan-

    Thank you for the update. We plan to use the IWR6843 in a surveillance application. Ideally, we want to maximize both the FOV and the range, although we realize that some tradeoffs will be likely. Beam steering seems ideal, as scanning the area once every 400-500 mS is more than adequate. Our goal is as close as possible to a 100m people detection range and a 120-degree FOV. We realize that some reduction in the range may occur at extreme angles (+60 and -60 degrees). As people walk closer to the building, we could switch to a MIMO mode to get faster updates and 3D info.

    Any guidance would be much appreciated. The IWR6843 is an awesome part!

    Thanks-

    Mike Schell

  • 0 in reply to Mike Schell
    TI__Mastermind 37805 points

    Hi Mike,

    Even without TX Beamforming you should be able to get 50+ meters of range at a 120 degree FOV for people detection on the IWR6843. My concern here is that you might introduce too much sidelobe into your beam if you make the angle too large. I'm still generating a final estimate, but perhaps you could use beamsteering to increase your range at say 0-45 degrees, then do a chirp without beamsteering to detect objects at the edges of the field of view. How does that sound?

    Best,

    Nate

  • 0 in reply to Nathan Block
    TI__Mastermind 37805 points

    Hi Mike,

    I've consulted internally and 60 degrees should be fine so long as your virtual antennas are spaced no more than half a wavelength apart (as all the TI EVMs are).

    Best,

    Nate 

  • 0 in reply to Nathan Block
    Prodigy 170 points

    Hi Nathan-

    Thank you, this is excellent news! Will I have to modify the SDK to support more frames? I understand the current program is limited to 4. I am assuming six frames of 20-degree width to cover the total 120-degree FOV. Does this seem reasonable? 

    Sorry for the added questions.

    Best-

    Mike

  • 0 in reply to Mike Schell
    TI__Mastermind 37805 points

    Hi Mike,

    Yes you will need more frames, but I don't know if it will require modifying the SDK. I would start by modifying the chirp configuration, then the LRPD example project if needed. In general I think your plan for 6 chirps should be reasonable. I would also recommend that as you're designing your pipeline, you make sure to account for the limits in memory in the L3 cache, which is where your radar datacube will be stored.

    Best,

    Nate

  • 0 in reply to Nathan Block
    Prodigy 170 points

    Hi Nate-

    Thank you, this is very helpful. Yes, I understand how we might bump up against the limit of the available L3 cache. Thanks for the warning.

    Best-

    Mike