Abstract
This study discusses the problem of direction-of-arrival estimation (DOA) estimation for a monostatic multiple-input multiple-output (MIMO) radar system, and a novel sparse Bayesian learning (SBL) framework is presented. To lower the computational load, the matched array data is firstly compressed via reduced-dimension transformation. Then the problem of DOA estimation is linked to a sparse inverse problem. Finally, a forgotten factor-based root SBL algorithm is derived from hyperparameters learning, which can solve the off-grid problem by finding the roots of a polynomial. The proposed algorithm does not require the prior of the source number, and it can apply to the scenario with a small snapshot as well as coarse grid, thus it has a blind and robust characteristic. Numerical simulations verify the effectiveness of the proposed algorithm.
Highlights
The topic of multiple-input multiple-output (MIMO) radar system has aroused extensive attention in the past decade [1,2,3,4,5]
We focus on the uniform linear array (ULA)-based monostatic MIMO radar, which belong to the later
Compared with the traditional subspace-based methods, e.g. RC-multiple signal classification (MUSIC) method, RC-ESPRIT method and the off-grid sparse Bayesian learning (SBL) methods, the advantages of the proposed method are given as follows: (a) The proposed algorithm does not require the prior of sources number, which means the proposed method has a blind characteristic, while the subspace methods are not blind; (b) The proposed method can work with a small snapshot as well as coherent sources while the traditional subspace methods may invalid in such backgrounds. (c) The proposed method can achieve an accurate direction-of-arrival estimation (DOA) estimation while the traditional peak-searching methods can only find the on-grid solutions. (d) The proposed require fewer iteration steps than the existing off-grid SBL methods, and it has better estimation accuracy than the existing off-grid SBL methods
Summary
The topic of multiple-input multiple-output (MIMO) radar system has aroused extensive attention in the past decade [1,2,3,4,5]. We focus on the uniform linear array (ULA)-based monostatic MIMO radar, which belong to the later. Unlike the angle estimation problem in a bistatic MIMO radar, only one-dimensional angle needs to be determined in a monostatic scenario. It has been shown that the ROGSBL method is more suitable for a coarse grid, and it has more accurate DOA estimation performance than the OGSBI method. An improved DOA estimation algorithm is proposed for a ULA-based monostatic MIMO radar. A sparse inverse model is established that links the DOA estimation problem to an off-grid optimisation problem. An improved root-SBL algorithm is derived, in which a forgotten factor model is utilised to update the grid dynamically. Numerical experiments show that the proposed method provides a more accurate DOA estimation than the existing off-grid SBL methods. X ·n and X m· denote the nth column and the mth row of X , respectively
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
