Joint Angle Estimation and Array Calibration Using Eigenspace in Monostatic MIMO Radar

Liangliang Li,Tingting Fu,Zhiguang Han,Yongqin Yang,Mengxing Huang,Jianfeng Li

doi:10.1109/access.2020.2977221

Abstract

In this article, aiming at the presence of the unknown gain-phase uncertainties in monostatic multiple-input multiple-output (MIMO) radar, an eigenspace based algorithm for joint parameter estimation is proposed to achieve angle and array calibration. The initial estimation of the direction of arrival (DOA) can be obtained from the signal subspace achieved by the eigenvalue decomposition (EVD) of covariance matrix, and then an improved multiple signal classification (MUSIC)-based cost function is established for achieving more accurate estimation of DOA, in which a local searching is only required because of the initial estimation value. Eventually, with the help of the estimated DOAs and the noise subspace, the error vector of array gain-phase can be achieved. The proposed method is not only superior to the classical subspace-based algorithms in terms of angle and array gain-phase error estimation, such as MUSIC-like method, especially for closely-spaced targets, but also can be suitable for non-uniform linear arrays (non-ULA). What's more, the computational complexity of the proposed method can be considerably reduced since only a local searching is needed. Multiple simulation experiments are carried out to illustrate the performance of the proposed scheme.

Highlights

Multiple-input multiple-output (MIMO) radars simultaneously transmit orthogonal waveforms via exploiting multiple sensors and use an antenna array to receive the signals reflected by multiple targets
The angle estimation of the sources, or direction of arrival (DOA) estimation is especially crucial for MIMO radar, which has been an attractive topic over the past decades [6]–[8]
Eq (19), the improved method in this paper shows advantage in estimating angle over conventional multiple signal classification (MUSIC)-like method in the case of closely-spaced targets, while achieves a slight performance improvement in the general case, the corresponding experiments will be carried out in the simulation section to verify the correctness of the above inference

Summary

INTRODUCTION

Multiple-input multiple-output (MIMO) radars simultaneously transmit orthogonal waveforms via exploiting multiple sensors and use an antenna array to receive the signals reflected by multiple targets. L. Li et al.: Joint Angle Estimation and Array Calibration Using Eigenspace in Monostatic MIMO Radar vector. An ESPRIT-like algorithm was further presented in MIMO radar [29], which can simultaneously estimate the angle and gain-phase error. A high-order decomposition based method was proposed in [33] which can obtain the estimated angle value without the influence of gain-phase errors combined with the dot division, but it requires the array to be non-linear. In this paper, based on eigenspace, an joint angle and array calibration method for monostatic MIMO radar under the condition of unknown gain-phase error is proposed. In contrast to conventional subspaces based methods, the proposed algorithm performance can be improved in terms of joint estimation of angle and array gainphase, especially for closely-spaced targets. Re(.) and min(.) are to get the real part and minimum element from a vector, respectively

DATA MODELING

COMPLEXITY ANALYSIS AND ADVANTAGES SUMMARY

SIMULATION RESULTS

CONCLUSION