Abstract
This paper presents a method for designing transmit beampattern in Four Dimensional (4D) imaging automotive Multiple-Input Multiple-Output (MIMO) radars, employing the distance between the designed and desired beampatterns as the design metric. Utilizing the ℓ p -norm criteria, we consider a broader range of p values, specifically for p ≥ 2 and 0 < p ≤ 1, to enhance the optimization framework. The optimization problem formulated under this criteria is efficiently solved using the Block Successive Upper-Bound Minimization (BSUM) technique for discrete and continuous phase constraints. Our analysis verifies the convergence of the objective function and confirms the solution’s convergence, thereby establishing a new stopping criterion for this optimization process. Furthermore, we demonstrate that our proposed method outperforms the commonly used omnidirectional beampattern across various automotive scenarios, highlighting its superior adaptability and utility in multiple applications. Additionally, our methods demonstrate good performance and computational efficiency, making them suitable for real-time 4D imaging automotive BSUM radar applications.
Funding Information
  • Luxembourg National Research Fund (FNR) CORE Metacognitive Radar for Emerging Sensing Applications (METSA) Project (C22/IS/17391632)

This publication has 40 references indexed in Scilit: