Sparsity-driven moving target detection in distributed multistatic FMCW radars

We investigate target detection by employing joint sparse signal modeling and recovery techniques for widely distributed multistatic Frequency Modulated Continuous Wave (FMCW) radar systems using chirp modulation. The sparse domain considered is the space-velocity domain. While a vast literature on applying joint sparse modeling and recovery to distributed multistatic radars exists, the proposed models are often not scalable in terms of complexity and required processing power for high resolution in the space-velocity domain. Furthermore, while sparse modeling studies for multistatic FMCW radar only consider static targets, we introduce a complete and generic sparse modeling in the space-velocity domain to jointly represent baseband radar signals. From this model, simplifications are introduced and a trade-off is drawn between the gain in computational complexity and the impact on the accuracy due to added model mis-match. A low-complexity adaptation of the Matching Pursuit algorithm relying on this simplified FMCW model is designed, and an iterative algorithm to compensate for the errors caused by these model simplifications is introduced. The effectiveness of the proposed method is demonstrated through Monte-Carlo simulations. The method achieves a fast estimation of moving target’s parameters on dense grids compared to previous sparsity-driven works on multistatic FMCW radars.