AWR2944EVM: Removal of static clutter

Hi Lokesh,

Please find my responses as below:

1. Can you let me know which version of the SDK you are currently using?
2. For the DDM mode, currently clutter removal option has not been added and is also not performed in the chain. The simplest way for you to implement something like that would be to filter the points based on velocity in the point cloud (i.e you can drop points with zero velocity).
3. For the TDM mode, the clutter removal feature should be functional. There is a doppler DC correction which should remove all signals that correspond to zero velocity. Can you let me know if this switch is happening or not when you enable the clutter removal option via the CLI by adding a breakpoint in the code as shown below? Can you also describe what observation made you arrive at the thought the clutter removal option might not be working? Is it that the point cloud spread did not change or point cloud with respect to the stationary points alone did not change etc.?
   1. 
4. CFAR does exclusively not care about whether the points are stationary or not. It depends on the sub-spectrum of the FFT that would be passed into the CFAR window. Having said that, you can reduce the number of stationary points (just like any other non-stationary points) by adjusting the CFAR threshold.

Regards,

Kaushik

Hello Kaushik,

Hello Kaushik,

Hello Kaushik,

Please find the screen shots of the plots with and without clutter removal option enabled. You can see there is no impact on the cloud points at all. I have also pasted the console messages, that confirms the enabling/disabling of clutter removal option.

Console messages for Clutter Removal On:

<<

mmwDemo:/>sensorStop
Ignored: Sensor is already stopped
Done

mmwDemo:/>flushCfg
Done

mmwDemo:/>dfeDataOutputMode 1
Done

mmwDemo:/>channelCfg 15 5 0
Done

mmwDemo:/>adcCfg 2 0
Done

mmwDemo:/>adcbufCfg -1 1 1 1 1
Done

mmwDemo:/>lowPower 0 0
Done

mmwDemo:/>profileCfg 0 77 267 7 57.14 0 0 70 1 656 13349 0 0 30
Done

mmwDemo:/>chirpCfg 0 0 0 0 0 0 0 1
Done

mmwDemo:/>chirpCfg 1 1 0 0 0 0 0 4
Done

mmwDemo:/>frameCfg 0 1 16 0 656 100 1 0
Done

mmwDemo:/>lowPower 0 0
Done

mmwDemo:/>guiMonitor -1 1 1 0 0 0 1
Done

mmwDemo:/>cfarCfg -1 0 2 8 4 3 0 15 1
Done

mmwDemo:/>cfarCfg -1 1 0 4 2 3 1 15 1
Done

mmwDemo:/>multiObjBeamForming -1 0 0.5
Done

mmwDemo:/>calibDcRangeSig -1 0 -5 8 256
Done

mmwDemo:/>clutterRemoval -1 1
Done

mmwDemo:/>compRangeBiasAndRxChanPhase 0.0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
Done

mmwDemo:/>measureRangeBiasAndRxChanPhase 0 1.5 0.2
Done

mmwDemo:/>aoaFovCfg -1 -90 90 -90 90
Done

mmwDemo:/>cfarFovCfg -1 0 0 22.88
Done

mmwDemo:/>cfarFovCfg -1 1 -1 1.00
Done

mmwDemo:/>extendedMaxVelocity -1 0
Done

mmwDemo:/>calibData 0 0 0
Done

mmwDemo:/>spreadSpectrumConfig 1 1 0.25 1 0 0 0 0 0 0 0 0
Done

mmwDemo:/>sensorStart
Done

>>

Console message for Clutter Removal Off:

<<

mmwDemo:/>sensorStop
Ignored: Sensor is already stopped
Done

mmwDemo:/>flushCfg
Done

mmwDemo:/>dfeDataOutputMode 1
Done

mmwDemo:/>channelCfg 15 5 0
Done

mmwDemo:/>adcCfg 2 0
Done

mmwDemo:/>adcbufCfg -1 1 1 1 1
Done

mmwDemo:/>lowPower 0 0
Done

mmwDemo:/>profileCfg 0 77 267 7 57.14 0 0 70 1 656 13349 0 0 30
Done

mmwDemo:/>chirpCfg 0 0 0 0 0 0 0 1
Done

mmwDemo:/>chirpCfg 1 1 0 0 0 0 0 4
Done

mmwDemo:/>frameCfg 0 1 16 0 656 100 1 0
Done

mmwDemo:/>lowPower 0 0
Done

mmwDemo:/>guiMonitor -1 1 1 0 0 0 1
Done

mmwDemo:/>cfarCfg -1 0 2 8 4 3 0 15 1
Done

mmwDemo:/>cfarCfg -1 1 0 4 2 3 1 15 1
Done

mmwDemo:/>multiObjBeamForming -1 0 0.5
Done

mmwDemo:/>calibDcRangeSig -1 0 -5 8 256
Done

mmwDemo:/>clutterRemoval -1 0
Done

mmwDemo:/>compRangeBiasAndRxChanPhase 0.0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
Done

mmwDemo:/>measureRangeBiasAndRxChanPhase 0 1.5 0.2
Done

mmwDemo:/>aoaFovCfg -1 -90 90 -90 90
Done

mmwDemo:/>cfarFovCfg -1 0 0 22.88
Done

mmwDemo:/>cfarFovCfg -1 1 -1 1.00
Done

mmwDemo:/>extendedMaxVelocity -1 0
Done

mmwDemo:/>calibData 0 0 0
Done

mmwDemo:/>spreadSpectrumConfig 1 1 0.25 1 0 0 0 0 0 0 0 0
Done

mmwDemo:/>sensorStart
Done

>>

I will be able to check whether the code in dopplerprochwa.c executes as expected after you advise how can I build the datapath libraries for CCS debug.

Regards and Thank you

Lokesh

Hi Lokesh,

Thank you for the details. Please find my responses as below:

1. Thanks for the confirmation.
2. I understand that you have the sensor mounted and point cloud-based filtering might not be suitable. Can you go through the following thread that discusses the RANSAC algorithm which is used in detecting stationary objects in a scene?
   AWR2944: How to understand ransac algorithm in AWR2944's demo "lab0015_high_end_corner_radar" - Sensors forum - Sensors - TI E2E support forums
   You can implement the same into your demo.
3. You can rebuild the libs using the command "gmake lib" in the respective DPU folder where the makefile is present. This should give you the debug symbols which can be used to place breakpoints. Alternatively, you can also place the breakpoints in the respective disassembly code.

Regards,

Kaushik

Hello Kaushik,

It is also the same at the place, you had highlighted in a picture above.

Regards and Thank you

Lokesh

Hello Kaushik,

Thank you for the reply. Please find my response,

1. I get this error reported in CCS console, when I do as suggested - 

Error: DPM Report 4 received with  error:-30204 arg0:0x64 arg1:0x1024eb30
[C66xx_DSP] Error: DPM Report 4 received with error:-30204 arg0:0x64 arg1:0x837a04
[Cortex_R5_0] ASSERT: 55.558961s: ./mss/mss_main.c:MmwDemo_DPC_ObjectDetection_reportFxn:2508: 0 failed !!!

2. I tried to trace the passing of the config info from mss_cli to it's use in doppleprochwa.c, the procedure is quite complex and passed thru handles and function pointers. If you could provide some design flow for passing of the config data between mss and dss, I could try and find the bug.

Regards and Thank you

Lokesh