RadarSimPy

A Radar Simulator for Python

RadarSimPy is a powerful and versatile Python-based Radar Simulator that models radar transceivers and simulates baseband data from point targets and 3D models. Its signal processing tools offer range/Doppler processing, direction of arrival estimation, and beamforming using various cutting-edge techniques, and you can even characterize radar detection using Swerling’s models. Whether you’re a beginner or an advanced user, RadarSimPy is the perfect tool for anyone looking to develop new radar technologies or expand their knowledge of radar systems.

Key Features

• Radar Modeling

  • Radar transceiver modeling
  • Arbitrary waveform
  • Phase noise
  • Phase/amplitude modulation
  • Fast-time/slow-time modulation

• Simulation

  • Simulation of radar baseband data from point targets
  • Simulation of interference
  • Simulation of radar baseband data from 3D modeled objects/environment (#raytracing)
  • Simulation of target's RCS (#raytracing)
  • Simulation of LiDAR point cloud from 3D modeled objects/environment (#raytracing)

• Signal Processing

  • Range/Doppler processing
  • Direction of arrival (DoA) estimation
    • MUltiple SIgnal Classification (MUSIC) DoA estimations for a uniform linear array (ULA)
    • Root-MUSIC DoA estimation for a ULA
    • Estimation of Signal Parameters via Rational Invariance Techniques (ESPRIT) DoA estimation for a ULA
    • Iterative Adaptive Approach (IAA) for amplitude and phase estimation
  • Beamformer
    • Capon beamformer
    • Bartlett beamformer
  • Constant false alarm rate (CFAR)
    • 1D/2D cell-averaging CFAR (CA-CFAR)
    • 1D/2D ordered-statistic CFAR (OS-CFAR)

• Characterization

  • Radar detection characteristics based on Swerling's models

Dependence

• numpy
• scipy
• pymeshlab (preferred) or meshio
• Visual C++ Runtime (Windows)

Installation

Download the pre-built module, and put the radarsimpy folder within your project folder as shown below:

---

• Windows

  • your_project.py
  • your_project.ipynb
  • radarsimpy
    • __init__.py
    • radarsimcpp.dll
    • simulator.xxx.pyd
    • rt.xxx.pyd
    • radar.py
    • processing.py
    • ...

---

• Linux

  • your_project.py
  • your_project.ipynb
  • radarsimpy
    • __init__.py
    • libradarsimcpp.so
    • simulator.xxx.so
    • rt.xxx.so
    • radar.py
    • processing.py
    • ...

---

• MacOS

  • your_project.py
  • your_project.ipynb
  • radarsimpy
    • __init__.py
    • libradarsimcpp.dylib
    • simulator.xxx.so
    • rt.xxx.so
    • radar.py
    • processing.py
    • ...

---

Acceleration

This module supports CPU/GPU parallelization. CPU parallelization is implemented through OpenMP. GPU parallelization (CUDA) has been added since v6.0.0.

	CPU (x86_64)	CPU (ARM64)	GPU (CUDA)
Windows	✔️	❌️	✔️
Linux	✔️	❌️	✔️
MacOS	✔️	❌️	❌️

Coordinate Systems

• Scene Coordinate

  • axis (m): [x, y, z]
  • phi (deg): angle on the x-y plane. 0 deg is the positive x-axis, 90 deg is the positive y-axis
  • theta (deg): angle on the z-x plane. 0 deg is the positive z-axis, 90 deg is the x-y plane
  • azimuth (deg): azimuth -90 ~ 90 deg equal to phi -90 ~ 90 deg
  • elevation (deg): elevation -90 ~ 90 deg equal to theta 180 ~ 0 deg

• Object's Local Coordinate

  • axis (m): [x, y, z]
  • yaw (deg): rotation along the z-axis. Positive yaw rotates the object from the positive x-axis to the positive y-axis
  • pitch (deg): rotation along the y-axis. Positive pitch rotates the object from the positive x-axis to the positive z-axis
  • roll (deg): rotation along the x-axis. Positive roll rotates the object from the positive z-axis to the negative y-axis
  • origin (m): [x, y, z]
  • rotation (deg): [yaw, pitch, roll]
  • rotation rate (deg/s): [yaw rate, pitch rate, roll rate]

Usage Examples

The source files of these Jupyter notebooks are available here.

• Radar modeling and point target simulation

  • Doppler radar
  • FMCW radar
  • TDM MIMO FMCW radar
  • PMCW radar
  • Arbitrary waveform
  • Phase noise
  • CFAR
  • DoA estimation
  • Interference

• Radar modeling and 3D scene simulation with raytracing

  • Imaging radar
  • FMCW radar with a corner reflector
  • FMCW radar with a plate
  • FMCW radar with a car
  • Doppler of a turbine
  • Micro-Doppler
  • Multi-path effect

• 3D modeled target's RCS simulation

  • Corner reflector RCS
  • Plate RCS
  • Car RCS

• LiDAR point cloud

  • LIDAR point cloud

• Receiver characterization

  • Receiver operating characteristic (ROC)

Build

Building radarsimpy requires to access the source code of radarsimcpp. If you don't have access to radarsimcpp, please use the pre-built module.

• Windows CPU

build_win.bat

• Windows CUDA

build_win_cuda.bat

• Linux CPU

./build_linux.sh

• Linux CUDA

./build_linux_cuda.sh

API Reference

Please check the Documentation