Real-Time MIMO-SAR Signal Processing System on Multi-Core DSP

Multiple-input, multiple-output Synthetic Aperture Radar (MIMO-SAR) has emerged as a pivotal research direction in radar technology due to its advantages of high resolution, wide-swath imaging, robust anti-jamming capability, and system flexibility. However, the multi-antenna architecture and multichannel data processing significantly increase algorithmic complexity and computational resource demands, while waveform orthogonality challenges further limit its practical applications. To address the challenges of waveform orthogonality and realtime processing in MIMO-SAR systems, this paper proposes a novel multi-core DSP framework. Key innovations include: (1) A SYS/BIOS-driven master-slave architecture achieving 82% resource utilization via MessageQ-based synchronization; (2) An elevation DBF algorithm optimized by LUP decomposition, reducing matrix operation time by 55%; (3) TCP/IP-based LFM waveform transmission balancing orthogonality and throughput. Experimental results demonstrate a 75% processing time reduction compared to single-core DSP baselines, validated through point and scene target imaging.