Research On Target Parameter Estimation Of Frequency-controlled Array MIMO Radar

Target parameter estimation is an important issue for radar system applications.MIMO(Multiple Input Multiple Output)radars,which use array antenna structures at the transmitters and receivers,can achieve high-resolution target parameter estimation by transmitting multiple orthogonal waveforms.Therefore,they have received widespread attention and become a hot spot in radar system design and research.MIMO radar is essentially a generalization of traditional phased array radar,which uses waveform diversity to achieve equivalent large-scale virtual arrays with fewer array elements,thereby improving the accuracy of target direction estimation.However,phased array radars cannot achieve range-azimuth-dependent beampatterns,which limits the ability of MIMO radars to resolve multiple targets at different range bins at the same azimuth bin.The range-independent pattern also restricts the application of MIMO radar in easy-to-intercept and easy-to-monitor scenarios.The Frequency Diverse Array(FDA)adds a frequency increment related to the position of the array element to the transmitted signal to achieve a range-azimuthdependent beampattern.These advantages have made frequency-divers arrays a hot topic of array radar research in recent years.Applying a frequency-diverse array to a MIMO radar is called a frequency-diverse array MIMO(Frequency Diverse ArrayMultiple Input Multiple Output,FDA-MIMO)radar.In the FDA-MIMO radar,the transmitting array can realize a large-scale virtual frequency-diverse array by transmitting multiple orthogonal signals with frequency increments,thereby achieving high-precision range-azimuth estimation.However,the frequency-diverse array also increases the complexity of signal processing in FDA-MIMO radar.How to use the characteristics of a frequency-diverse array to effectively perform joint range-azimuth estimation or joint range-azimuth-velocity estimation is of great significance to the application of FDA-MIMO radar.This paper studies the target parameter estimation problem for FDA-MIMO radar,and proposes the joint range-azimuth estimation method and the joint range-azimuthvelocity estimation method for FDA-MIMO radar.The performances of the FDAMIMO radar are analyzed,both theoretically and numerically.The main contributions of the paper are as follows:(1)A two-dimensional atomic norm-based joint range-azimuth joint estimation algorithm for FDA-MIMO radar is proposed.This algorithm uses the structural characteristics of received signals of FDA-MIMO radar to define a twodimensional atomic norm,and transforms the joint range-azimuth estimation problem into a low-rank matrix reconstruction problem.This method avoids the off-grid problem in the sparse reconstruction and improves the accuracy of parameter estimation.Simulation experiments show that this method can effectively achieve high-precision joint range-azimuth estimation of FDAMIMO radar.(2)The Cramér-Rao Lower Bound(CRLB)on the joint range-azimuth joint estimation of the FDA-MIMO radar is theoretically analyzed.The expression of the CRLB on the joint range-azimuth estimation is derived,and the relationship between the parameter estimation performance and the frequency increment of the transmitted signal is established.Based on this analysis,a CRLB-based optimization method to choose the incremental frequency of transmitted signals in the FDA-MIMO radar is also proposed.(3)For the moving target model,two kinds of joint range-azimuth-velocity estimation methods for FDA-MIMO radar are proposed.The first one is developed based on the slices of two-dimensional atomic norm.The idea is to decompose the three-dimensional parameter estimation problem into multiple two-dimensional parameter estimation problems,and then realize the threedimensional parameter estimation through a pairing algorithm.The second one is a tensor-based joint range-azimuth-velocity estimation method that uses a parallel factor algorithm to directly solve the three-dimensional parameter estimation problem.Simulation experiments analyze and verify the effectiveness and accuracy of the proposed algorithms.