Joint DFT-ESPRIT Estimation for TOA and DOA in Vehicle FMCW Radars

Abstract

This letter proposes a joint discrete Fourier transform (DFT)-estimation of signal parameters via rotational invariance techniques (ESPRIT) estimator for time-of-arrival (TOA) and direction-of-arrival (DOA) in vehicle frequency-modulated continuous-wave (FMCW) radars. Since the vehicle FMCW radar should recognize vehicles in the side/rear area when the driver initiates a lane change, the estimation of the joint TOA/DOA between the radar and targets is an important issue for solving complicated location tasks. However, conventional joint estimation methods such as 2D-ESPRIT and 2D-multiple signal classification (MUSIC) cannot be adopted for real-time implementation due to their high computational loads. To satisfy the required accuracy specifications and reduce complexity compared with the conventional estimator, we propose a low-complexity joint TOA and DOA estimator that uses the combined DFT-ESPRIT algorithm for FMCW radars. The performance of the proposed estimation in multitarget environments was derived and compared with the Monte Carlo simulation results. The root-mean-square error (RMSE) of the proposed method was compared with that of 2D-ESPRIT with various parameters. To verify the performance of the proposed combination method, we implemented the FMCW radar and verified its performance in an anechoic chamber environment.

An Equivalent Method of Position Error Caused by the Array Antenna Deformation

“…According to the subspace theory, the response matrix A of the array satisfies the relationship with the signal subspace E, wherein: E = AT (14) As the array receives only one signal source, the response matrix A degrades into the steering vector a:…”

“…The rotation invariance of the subarray generated by the displacement invariance of the subarray is used to solve the direction of the signal according to the rotation phase between the subfields of the subarray. Compared with other high-resolution DOA estimation algorithms, this algorithm does not need angle search, and the estimation result can be directly calculated [14][15][16].…”

Joint Range and Angle Estimation of FMCW MIMO Radar

“…Either with the full Rx array or with the MIMO virtual array, we first form a range-angle matrix, where the angle dimension may extend to the size of the virtual array in case of the MIMO system. Then, to get range and angle estimates of each target, we may apply two-dimensional (2D) frequency estimation algorithms on the range-angle matrix, such as the 2D fast Fourier transformation (2D-FFT) [13], joint angle-frequency estimation (JAFE) [16], 2D multiple signal classification (2D-MUSIC) [14], or the joint discrete Fourier transform (DFT)-estimation of signal parameters via rotational invariance techniques (ESPRIT) [17].…”

Reduced-dimensional MUSIC Algorithm for Joint Angle and Delay Estimation based on L2 Norm Constraint in Multipath Environment

“…Maximum likelihood (ML)-based methods [1], ESPRIT based methods [2,3] and MUSIC-based methods [4][5][6] are the common estimation methods of great interest. Particularly, the ML estimator (MLE) is an optimal technique.…”

“…Although some iterative algorithms have been developed to reduce the complexity, such as the Gauss Newton-based method in [1], they require accurate initial guesses to avoid the local convergence. ESPRIT-based algorithms [2,3] usually have closed-form solutions and low complexity. In [2], the JADE-ESPRIT algorithm is proposed which utilizes the rotational invariance property of the array, but its performance is significantly worse compared with the Cramer-Rao bound (CRB), especially at low signal-to-noise ratio (SNR).…”