GPU acceleration algorithm for parallelized pulse compression in inverse synthetic aperture radar

• Main Authors: Mo, Junxian, Chen, Jinfan, Xie, Hanfeng, Zheng, Zixiang, Zhang, Yue
• Format: Conference Proceeding
• Language: English
• Published: SPIE 13.06.2024
• ISBN: 151068042X; 9781510680425
• ISSN: 0277-786X
• DOI: 10.1117/12.3033809

The current development of ISAR imaging is towards higher resolution, using a larger transmission signal bandwidth, resulting in a huge scale of echo data. Pulse compression is an important component of ISAR imaging, which is the prerequisite and foundation of imaging. Under the direct acquisition method of intermediate frequency, the sampling data scale of high bandwidth radar is very large, and traditional DSP processing platforms and serial processing methods are difficult to meet real-time imaging requirements. Therefore, this article proposes a parallelized pulse compression algorithm, which runs on a graphics processing unit (GPU) and utilizes the similarity of radar echo processing methods to design a parallelized digital de skewing algorithm while completing pulse compression. This method can greatly improve processing efficiency, quickly obtain a one-dimensional range profile sequence of the target, and effectively reduce the amount of processed data through target area extraction, which can better achieve real-time ISAR imaging. The measured data validation shows that the efficiency of this method can be improved by more than 9 times, which verifies the advantages and effectiveness of the algorithm.