Reduced-dimensional MUSIC Algorithm for Frequency Diverse Array in MIMO Radar System

The paper proposes a reduced-dimensional multiple signal classification (RD-MUSIC) algorithm for frequency diverse array (FDA) multi-input multi-output (MIMO) radar. The proposed method firstly constructs a constrained minimization problem based on the two-dimensional (2D) MUSIC spectrum function. Then, we reduce the dimension of the minimization problem by calculating the partial derivative with respect to the range parameter, which results in the minimization problem only about the direction of arrival (DOA) parameter. Finally, we substitute the DOA estimates into the steering vector and directly obtain the closed-form solutions of range estimates based on the least squares (LS) method. The proposed algorithm requires only one 1D peak search for DOA parameters, which greatly reduces the computational complexity. In addition, 1D peak searches can be used for the range estimation as well, achieving better estimation accuracy at the cost of the increased complexity compared with the LS solutions. Moreover, the proposed algorithm can be applied to not only uniform linear array (ULA) but also sparse array, demonstrating high versatility. Numerical results show that the proposed RD-MUSIC algorithm is effectiveness regardless of the array configuration and that it can outperform the conventional algorithms including atomic norm minimization (ANM) based method and MUSIC based method.