Loading...

Angle-incremental range estimation for FDA-MIMO radar via hybrid sparse learning

Karbasi, S. M ; Sharif University of Technology | 2022

118 Viewed
  1. Type of Document: Article
  2. DOI: 10.1016/j.dsp.2022.103749
  3. Publisher: Elsevier Inc , 2022
  4. Abstract:
  5. In this paper, a target parameter estimation problem is addressed for the recently emerging frequency diverse array multiple-input multiple-output (FDA-MIMO) radar system, utilizing sparse learning. The scene is modeled as a two dimensional (2D) angle-incremental range grid. To solve the resulting sparse problem, the recently proposed user-parameter free algorithms including block sparse learning via iterative minimization (BSLIM), iterative adaptive approach (IAA), sparse iterative covariance-based estimation (SPICE), likelihood-based estimation of sparse parameters (LIKES), and orthogonal matching pursuit (OMP) are applied which achieve excellent parameter estimation performance. However, these algorithms do not exhibit a uniform performance under different scenarios such as the number of available snapshots and the number of present targets. In this paper, we propose a hybrid estimation approach which utilizes the model order selection (MOS), and automatically selects the best algorithm for the scenario under test. The performance analyses show that the proposed hybrid algorithm presents higher accuracy in terms of root mean square error (RMSE), as compared with the discussed sparse algorithms, and converges to the Cramér-Rao lower bound (CRLB). © 2022 Elsevier Inc
  6. Keywords:
  7. CRLB ; FDA-MIMO ; Sparse recovery ; Adaptive algorithms ; Frequency estimation ; Iterative methods ; Learning algorithms ; MIMO radar ; MIMO systems ; Angle ; Crame-Rao lower bounds ; Cramer Rao lower bound ; Frequency diverse array multiple-input multiple-output ; Incremental range ; Multiple inputs ; Multiple outputs ; Multiple-input multiple-output radars ; Range estimation ; Mean square error
  8. Source: Digital Signal Processing: A Review Journal ; Volume 130 , 2022 ; 10512004 (ISSN)
  9. URL: https://www.sciencedirect.com/science/article/abs/pii/S1051200422003669