Parallel Computing Models with MATLAB for Radar Digital Twins using Multiple CUDA GPUs
Abstract
The article discusses the issues of parallel computational models of various elements of radar systems with MATLAB, for calculating on the multiple CUDA GPUs, and intended for use in digital twins of radars. At the beginning of the article, the principles of constructing of radars digital twins are briefly described. The main content of the article considers a number of computational problems solved by the authors and the corresponding computational MATLAB models, which can be used to create digital twins of radars: - antenna array radiation pattern calculation; - multi-beam MIMO antenna array radiation pattern calculation; - radar signal ambiguity function calculation; - calculation of a digital radio hologram (DRH) of the reflected radio signal; - radar image synthesis in the monostatic mapping SAR; - radar image synthesis in multi-frame mapping of a long strip; - radar image synthesis in autofocus mode SAR; - radar image synthesis in the moving target indication (MTI) mode SAR; - moving target detection using the algorithm with a Constant False Alarm Rate (CFAR); - radar image synthesis in monostatic MIMO-SAR. The ways of dividing the listed computing tasks into GPUs are presented. The calculation time reduce with multiple GPUs using is shown in comparison with calculations with one GPU and with the central processing unit (CPU).
Fields of Study & Subjects
Advanced SAR Imaging Techniques
Parallel Computing and Optimization Techniques
Radar Systems and Signal Processing
Publication details
- Journal / source
- Radioelectronics Nanosystems Information Technologies
- Volume
- 18
- Issue
- 1
- Pages
- 41–54
- Publication date
- 15 February 2026
- Language
- EN