25.02.2026 aktualisiert
FH
100 % verfügbar
Signal Processing Engineer
München, Deutschland
Weltweit
Phd Electrical EngineeringSkills
AlgorithmusAlgorithmus-EntwurfComputer VisionAtlassian ConfluenceAtlassian JiraAutomotive ElectronicsBeamformingC++KalibrierungCUDAComputerprogrammierungLinuxEmbedded SoftwareGitHubPython
- Radar & DSP: FMCW, SAR, CFAR detection, DoA / beamforming, MIMO, calibration, simulation, radar performance analysis
- Programming: Python, MATLAB, C/C++, HLSL, CUDA
- Tools & SW: Git/GitHub, Jira, Confluence, Doxygen, LaTeX, Linux, Visual Studio / VS Code
- Simulation: Unreal Engine, OptiX, CST
Erfahrung
Berufserfahrungmehr als 10 Jahre
Freelancer seit2026
Sprachen
DeutschMutterspracheEnglischverhandlungssicher
Projekthistorie
Algorithm development for automotive radar signal processing:
- Developed radar signal processing algorithms in Python, covering calibration, Direction of Arrival (DoA) estimation, and target detection
- Designed and maintained complete radar processing pipeline from raw ADC data to point cloud output, maturing algorithms from research-grade prototype to deployment-ready implementation
- Collaborated with embedded software engineers to optimize algorithm runtime for deployment on resource-limited embedded platforms
- Conducted lab measurements and data analyses to validate algorithm performance and identify optimization potential
- Evaluated radar detection accuracy and resolution against ground truth references, assessing angle estimation performance to identify and drive targeted algorithm improvements
Radar simulation for training data generation in SAR-based object detection:
- Led the radar simulation team, coordinating workstreams across algorithm, rendering and ML engineering
- Developed a high-fidelity SAR image simulator in Unreal Engine C++, incorporating physically-based radar waveform and scene interaction effects
- Implemented ray-tracing models in HLSL for radar reflectivity and clutter simulation, enabling realistic material and geometry modelling
- Investigated waveform-scene interaction effects through simulation to improve training data realism
- Configured, trained, and evaluated YOLO-based object detection models on synthetic SAR radar imagery
Algorithm development for series-production automotive radar sensors (FMCW/77 GHz, ARS540):
- Designed and optimized radar raw-data processing algorithms including detection (CFAR), angle estimation (DoA/beamforming), and ambiguity resolution for mass-production radar sensors
- Prototyped algorithms in MATLAB and implemented them in C under strict real-time and memory constraints on dedicated radar hardware
- Coordinated with system and hardware teams to align algorithm design with sensor capabilities
- Planned and conducted vehicle/lab measurement campaigns; performed systematic failure analyses on field data to drive targeted improvements