Drastically cutting antenna element count and ADC power consumption by exploiting spatial target sparsity.
Because physical objects occupy only a tiny fraction of total 3D spatial voxels, the received radar signal is highly sparse in the angle-range-Doppler domain.
GPU-accelerated OMP reconstruction algorithms reconstruct complete 4D point clouds from 70% fewer physical antenna elements with zero loss of spatial fidelity.
Applying compressive sensing reduced baseband ADC and DSP power consumption by 54% while maintaining sub-decimeter target localization accuracy.