IWRL6432AOPEVM: IWRL6432AOP

Hi, I can parse the TLVs and understand which value is related to which data field. However I do not find anything relate to the 256 range bins data that how can I use it or what processing further is needed to convert it in a meaningful data. I have few queries as mentioned below.

1. If I want to control the operation of the radar with the GPIO interrupt of the host MCU what changes I should do and is there any GPIO pin available on the IWRL6432AOPEVM board using which I can test this?
2. Some times I noticed that after keeping an object in front of radar (when it was facing up to the sky in open)  at a distance near 1 meter I was getting this TLV - MMWDEMO_OUTPUT_EXT_MSG_DETECTED_POINTS   data and with few detected points, but All the Point Cloud data were having only SNR and NOISE non zero values and x, y, z, doppler, parameters were all Zero. What does it mean? If it is a valid data how do we get the distance of the detected point.
3. As per my understanding The Point Cloud data of this TLV - MMWDEMO_OUTPUT_EXT_MSG_DETECTED_POINTS gives the information regarding the location of detected point from sensor (in meters the value of x, y, z) in cartesian coordinate and it's radial velocity , snr and noise. In any case every detected point should be at specific coordinate(not on the origin) and at least one value of x, y, z will be non zero . Am I right? 

Thank You.

Hello, 

1 does not seem related to 2&3. Can you please create a separate post for that question?

Rampratap Chauhan said:

- MMWDEMO_OUTPUT_EXT_MSG_DETECTED_POINTS   data and with few detected points, but All the Point Cloud data were having only SNR and NOISE non zero values and x, y, z, doppler, parameters were all Zero.

These do not seem like valid detections. Have you modified the PresenceDetect.cfg configuration parameters or just using the default? Can you let me know the SDK version? Also, how often are you seeing these kind of points?

Rampratap Chauhan said:

As per my understanding The Point Cloud data of this TLV - MMWDEMO_OUTPUT_EXT_MSG_DETECTED_POINTS gives the information regarding the location of detected point from sensor (in meters the value of x, y, z) in cartesian coordinate and it's radial velocity , snr and noise. In any case every detected point should be at specific coordinate(not on the origin) and at least one value of x, y, z will be non zero . Am I right?

Your understanding is mostly correct. It is technically possible for there to be detected points at the origin due to the antenna coupling effect. But to avoid this we include the rangeSelCfg command in the configuration which controls the min and max distance that a detected point must have to be included in the output. Our default configurations should set the minimum value to 0.1m to remove detections caused by antenna coupling. Can you check the min value for rangeSelCfg in your configuration?

Best regards,

Josh

Hi, sure I will create a separate post for this.

1. I still did not get the answer of what I can do with the range bins data does it helps in improving the detection?
2. The 2nd query of getting all x, y, z value 0 in my previous question is resolved, I was getting incorrect data because of a parsing mistake.
3. I have not change any thing in the PresenceDetect.cfg apart from baud rate and the value of rangeSelCfg  is 0.1 to 4.
4. Now I am getting reasonably correct data for the objects in front of the sensor but I have noticed some false detection and some inaccurate value of point cloud  data as well. Like if I am getting 3 detection at 1 meter and 2 detection at .6 meter a confusion arises that which value should be considered correct ? What can we further do to improve the accuracy?