A simulator of mmWave radar that simulates the IF signal as if the radar is placed in a customized scene, for algorithm design and verification.
It supports configuration of:
- Number and layout of receiver array. It has several built-in layouts for Texas Instruments radars.
- Chirp configuration, such as slope, duration, frequency, number of chirps per frame, ADC sampling rate, etc.
- Object in the scene, represented as points. Support 3D model import from some pyTorch geometric datasets.
- Object motion.
- Signal-to-noise ratio.
The project contains algorithms for detecting the range, velocity, and angle-of-arrival of the objects and estimating a point cloud based on the simulation data.
Install Python (tested with 3.9) and install dependencies with pip install -r requirement.txt.
The demo-simple.py script demonstrates the basic functions of the tool. It sets up a scene consisting of a single point with a motion, and shows how the simulation data can be interpreted.
python demo-simple.py
The demo-pointcloud-eval.py script demonstrates how to import a human model from the FAUST dataset, get the simulation data, and estimate a point cloud based using the data.
python demo-pointcloud-eval.py
If you find this work useful, please consider cite:
@misc{Cui23,
doi = {10.48550/ARXIV.2301.13553},
url = {https://arxiv.org/abs/2301.13553},
author = {Cui, Han and Wu, Jiacheng and Dahnoun, Naim},
title = {Millimetre-wave Radar for Low-Cost 3D Imaging: A Performance Study},
publisher = {arXiv},
year = {2023}
}