• Resolved

IWR6843ISK-ODS: rangeAzimuthHeatMap>

Prodigy 160 points

Replies: 6

Views: 56

Part Number: IWR6843ISK-ODS

Hello everyone,
I recorded a data file from IWR6843ISK-ODS containing both rangeAzimuthHeatMap and rangeDopplerHeatMap. I attached both the data file and the first frame of rangeAzimuthHeatMap for the first column.
The first frame of rangeAzimuthHeatMap is a complex matrix with 12 columns, i.e. the number of virtual antennas (3x4), and 256 rows, i.e. the number of range bins. 
Before starting with the AoA computation I would make a first simple test by plotting just one column that I think is the range detected by one virtual antenna. Therefore, I expected to see something similar to a range FFT output but this is not the case. Thereafter, maybe I do not understand what is the rangeAzimuthHeatMap output of the device. Please can you clarify this aspect?
Sincerely,
Emanuele

xWR68xx_AOP_processed_stream_2020_07_30T06_12_59_395.datra_heatmapFirstAntenna.txt

  • HI, Emanuele:

    The range Azimuth heatmap data that transfer from EVM is just the zero Doppler output.  And the acutal range azimuth heatmap is calculated inside the visualizer on PC.   The range Doppler heatmap that transfer from EVM is the actual range Doppler heatmap. 

    Let me know whether it is making sense to you now.

    Best,

    Zigang

  • In reply to zigang Yang:

    Dear Zigang Yang,
    Thank you a lot for your answer. I understand what is the range Doppler heatmap but I have some problems with the range Azimuth heatmap. How is the zero Doppler output calculated? An additional source of confusion for me is this file:
    e2e.ti.com/.../6153.understand_5F00_OOB_5F00_output.pdf
    It is related to IWR6843ISK-ODS and it shows both the terms “Azimuth Static Heatmap” and “Azimuth/Elevation Static Heatmap”. What is the difference?
    Finally, with some details about my purpose, maybe you can suggest the best solution. I need to measure the phase of the received signal for detecting the vital sign.
    Usually, this is a trivial operation starting from the time domain data, computing the range profile and then extracting the phase with the arctangent operator. Unfortunately, I do not have a DCA1000 board yet (for receiving the raw time domain data), so my idea was to start directly from the complex range profile. I had understood that EVM output is a complex range profile for each virtual antenna and consequently I could extract the phase history for each virtual antenna but from your answer, it seems not the case.
    Have you suggestions for extracting the target phase for measuring small movements starting from the UART data provided by the mmWave Demo Visualizer?
    Sincerely,
    Emanuele

  • In reply to Emanuele Cardillo:

    HI, Emanuele:

    The output in mmwave OOB demo does not provide any phase information for range-Doppler heat-map.  As I mentioned, the range-azimuth heatmap is generated by zero Doppler value.  Zero Doppler value is complex, it is computed as sum of all chirps for each range bin per antenna.  But if you need phase information for different  chirp, then you can not use it.   

    Using the zero Doppler value, you can get range-azimuth heatmap, i.e., generate the azimuth angle spectrum for each range.   You can also generate range-elevation heatmap.  

    Best,

    Zigang

  • In reply to zigang Yang:

    Hello Zigang,
    thank you for your kind answer, now I understand that it is not possible to obtain phase information from this board without a DCA1000 board (just a quick question. My setup is composed by: IWR6843ISK-ODS + MMWAVEICBOOST + proper external 5V power supply. By adding a DCA1000, will I need an additional external 5V power supply like that of MMWAVEICBOOST or all these three board could be powered by one 5V power supply?).
    Concerning the zero Doppler, there is something else that I do not understand. It is computed as sum of all chirps for each range bin per antenna. Does this mean that is is something similar to the range profile? So is it equal to 12 range profiles each one detected by each virtual antenna? I suppose that the answer is again negative. Indeed, I tried to plot directly only this data for only one virtual antenna but it seems like a time domain signal. Do I need to perform an fft on these data for obtaining the range profile? For me the range profile for just one virtual antenna is an amplitude vs range plot (or equivalently amplitude vs frequency plot).
    Maybe, for obtaining the range-azimuth heatmap, first I need to perform a range fft. Then, I have to take into consideration only the data from the azimuth virtual antennas and performing an additional fft over the antennas for obtaining the range-azimuth heatmap. In the same way, I can consider only the data from the elevation virtual antennas and performing an additional fft over these antennas for obtaining the range-elevation heatmap. Is this correct?
    Best Regards,
    Emanuele

  • In reply to Emanuele Cardillo:

    HI, Emanuele:

    It is possible to use one 5V power supply for the 3 board (IWR6843ISK-ODS + MMWAVEICBOOST + DCA1000).  The DCA1000 board has a switch to choose its own power supply or feed from ICbooster board. 

    The zero Doppler should be complex data.  If you take the amplitude of the complex number, it should looks similar as range profile.  If you are interested in the details, please dig into the source code to find the details. 

    After you get the raw data captured with DCA1000 board, you should able to get the phase change for each range bins.   The range-angle heatmap should not really matter to you.

    Best,

    Zigang 

  • In reply to zigang Yang:

    Dear Zigang,

    thank you for your answer.

    Emanuele