Introduction and demos of the mmWaveMatlab project.

mmWaveMatlab is a Matlab codebase for millimeter wave radar signal processing and human sensing. This document will introduce some of the main features of mmWaveMatlab and show the corresponding results.

Provide basic range, velocity, and angle measurements.

Perform Range FFT to obtain range spectrums.

1. Perform Doppler FFT to obtain RD maps.

3. Compute micro-Doppler spectrums. Support simultaneous observation of multiple targets through range/RA extraction.

1. Configure the antenna array and perform angle estimation algorithms (FFT or digital beamforming) to obtain RA maps. Digital beamforming supports CBF and Capon algorithms and provides signal reconstruction.

2. Perform 2D angle estimation and accumulate signal strength for human imaging.

Provide point cloud-related functions.

Perform 1D/2D CFAR on range spectrum, RD Map, and RA map to generate point clouds.

$~~$

Generate 4D point clouds through approaches of 4D FFT (RD point cloud -> 2D Angle FFT) or 2-Pass DBF (RA point cloud -> elevation estimation & signal reconstruction -> Doppler FFT).

1. Cluster 2D/3D point clouds of different types through DBSCAN.

3. Distinguish nearby targets through GMM clustering.

Provide various human sensing functions.

A multi-person tracking solution with full process optimization involving detection enhancement, interference suppression, continuity maintenance, and trajectory correction. It can be used in conjunction with IMUs to identify users [1].

$~~~~$

$~~~~$

Perform respiration monitoring on static targets. Support simultaneous observation of multiple targets through range/RA extraction.

Welcome to our research on mmWave radar, and feel free to contact us.

[1] "PmTrack: Enabling personalized mmWave-based human tracking," ACM IMWUT, 2023. (Multi-person tracking and identification)

[2] "Application-oriented privacy filter for mmWave radar," IEEE Communications Magazine, 2023. (Behavior privacy protection)

[3] "PosMonitor: Fine-grained sleep posture recognition with mmWave radar," IEEE Internet of Things Journal, 2024. (Sleep posture recognition)

[4] "VibCamera: mmWave and camera fusion for multi-point vibration monitoring", IEEE ICPADS, 2023. (Vibration measurement)