DESIGN AND IMPLEMENATION OF GNSS-INBSAR REAL-TIME PREPROCESSING SYSTEM BASED ON SOC

The GNSS-InBSAR system, based on navigation satellites, utilizes on-orbit satellites as transmitters and combines radar with navigation signals to achieve continuous deformation monitoring of large-scale scenes. Due to the weak signal-to-noise ratio, low resolution, and poor coherence of the reflected signals from the scene, traditional reflection signal processing algorithms are time-consuming and inefficient in data processing. This paper designs and implements a real-time preprocessing system for GNSS-InBSAR based on SoC. It proposes a time-frequency phase synchronization method based on direct waves and segmented pulse compression technology. The problem of insufficient signal processing resources for multi-frequency point and multi-channel signals is solved through FFT multiplexing technology. The CORDIC algorithm is used to calculate compensation phase, improving phase accuracy. The proposed methods are validated in MATLAB. Finally, the reflection signal pulse compression results and backprojection (BP) imaging performance were tested and validated. The test results indicate that the system can efficiently and stably perform synchronous acquisition and preprocessing of direct wave and echo signals from 3 frequency points and 8 channels. SAR images with high point density and resolution are obtained through azimuth accumulation based on the preprocessed data, meeting the accuracy requirements for deformation measurement of large-scale scenes.