Pythonic mmWave Toolbox for TI's IWR Radar Sensors

Introduction

This is a toolbox composed of Python scripts to interact with TI's evaluation module (BoosterPack) for IWRxx43 mmWave sensing devices. The toolbox provides easy access to particular OOB firmware versions, which are included in TI's mmWave SDKs and Industrial Toolboxes while focusing on data capturing and visualization with Python 3. Some elements of the toolbox can be considered being a Pythonic alternative to TI's mmWave Demo Visualizer.

Support

• IWR1443 ES2.0
  • Capture Demo (SDK 1.1.0.2)
  • mmWave SDK Demo (SDK 1.2.0.5)
  • High Accuracy 14xx (Industrial Toolbox 2.5.2)
  • 4K FFT Stitching (Industrial Toolbox 2.5.2)
• IWR1443 ES3.0
  • mmWave SDK Demo (SDK 2.1.0.4)
  • High Accuracy 14xx (Industrial Toolbox 4.1.0)
• IWR1843 ES1.0
  • mmWave SDK Demo (SDK 3.4.0.3)
• IWR6843 ES2.0
  • mmWave SDK Demo (SDK 3.4.0.3)

Advanced frame configuration with subframes is yet not supported.

Features

• 2D plots
  • range and noise profile
  • Doppler-range FFT heat map
  • azimuth-range FFT heat map
  • FFT of IF signals
• 3D plots
  • CFAR detected objects (point cloud)
  • simple CFAR clustering
• Data capture
  • range and noise profile with CFAR detected objects

3D plots are currently not implemented for IWR1843 ES1.0 and IWR6843 ES2.0.

Usage

The launcher of the toolbox is pymmw.py:

usage: pymmw.py [-h] [-c PORT] [-d PORT] [-f HANDLER] [-n]

arguments:
  -h, --help                            show this message and exit
  -c PORT, --control-port PORT          set port for control communication
  -d PORT, --data-port PORT             set port for data communication
  -f HANDLER, --force-handler HANDLER   force handler for data processing
  -n, --no-discovery                    no discovery for USB devices

In GNU/Linux, the launcher attempts to find and select serial ports of a valid USB device if no serial ports are provided.

Internals

1. The launcher
   • performs a connection test and resets the device (via the XDS110 debug probe).
   • observes the control port and tries to detect the firmware by its welcome message.
   • dynamically imports and initializes the handler (from /mss/*.py) for the firmware.
2. The handler
   • reads its configuration file (from /mss/*.cfg) once at startup.
   • observes the data port and starts applications (from /app/*.py) defined in _meta_ with their required arguments.
   • captures data from the data port, possibly performs data preprocessing, and pipes the data to the application processes.
3. Each application process
   • takes the data from stdin.
   • performs postprocessing, mostly for visualization or capturing purposes.
   • puts data, if implemented, to a file in /log.

Be patient when using Capture Demo for the "FFT of IF signals" application as the ADC data is first copied from ADC buffers to L3 memory and then published via UART. The update interval is about 10 seconds.

Configuration

The configuration files are pretty much like the original TI's profiles, except that they are JSON-formatted. Furthermore, all entries without a value (set to null) are attempted to be filled up with inferred values from the additional _settings_ block, which contains advanced settings for postprocessing (like range bias due to delays in the RF frontend) and a generic antenna configuration.

Make sure to have an appropriate configuration file with reasonable settings related to the firmware and handler, respectively, residing in /mss. Several configuration files with exemplary settings for different profiles and use cases are stored in /mss/cfg.

Dependencies

The toolbox works at least under GNU/Linux and Windows 10 with Python 3.8.5 - 3.8.7 if the following dependencies are met:

• pyserial (3.4 - 3.5)
• pyusb (1.0.2 - 1.1.0)
• matplotlib (3.3.1 - 3.3.3)
• numpy (1.19.1 - 1.19.4)
• scipy (1.5.2 - 1.5.4) - is only required for the application "azimuth-range FFT heat map" of the mmWave SDK Demo
• tiflash (1.2.9) - is only required for the application "FFT of IF signals" of the Capture Demo in conjunction with Texas Instruments’s Code Composer Studio (8.3.0) scripting interface to read ADC data from the L3 memory
• pyftdi (0.51.2 - 0.52.0) - is only required for the tool "reset via FTDI"
• XDS Emulation Software Package (8.3.0) - is only required for working with Windows

To make the tiflash module and the "FFT of IF signals" application work properly, an environment variable named CCS_PATH should point to the Code Composer Studio directory.

Troubleshooting

Permission Issue

Insufficient permissions to access the device: probe usb_discover The device has no langid (permission issue, no string descriptors supported or device error). Add the required permissions for the XDS110 (and FTDI) device in udev, e.g. in /etc/udev/rules.d/99-mmwave.rules:

# XDS110 (xWR EVM)
SUBSYSTEM=="usb",ENV{DEVTYPE}=="usb_device",ATTRS{idVendor}=="0451",ATTRS{idProduct}=="bef3",MODE:="0666"

# FTDI (DCA, ICBoost)
SUBSYSTEM=="usb",ENV{DEVTYPE}=="usb_device",ATTRS{idVendor}=="0451",ATTRS{idProduct}=="fd03",MODE:="0666"

KERNEL=="ttyACM[0-9]*",MODE:="0666"

No Handler

No handler for a supported device and firmware could be found: pymmw _read_ no handler found. If pymmw.py is not able to read the welcome message of the firmware for some reason, try to set the USB ports manually and disable the USB discovery:

pymmw.py -c /dev/ttyACM0 -d /dev/ttyACM1 --no-discovery

Use the NRST button to reset the device if the no-discovery option is activated.

Reference

If you find this toolbox or code useful, please consider citing our paper:

@inproceedings{constapel2019practical,
  title={A Practical Toolbox for Getting Started with mmWave FMCW Radar Sensors},
  author={Constapel, Manfred and Cimdins, Marco and Hellbr{\"u}ck, Horst},
  booktitle={Proceedings of the 4th KuVS/GI Expert Talk on Localization},
  year={2019}
}