Research On Multidimensional Parameters Estimation Algorithm Based On FMCW MIMO Radar

Frequency Modulated Continuous Wave(FMCW)Multiple Input Multiple Output(MIMO)radar is an advanced radar system that uses frequency-modulated continuous wave to measure target distance,velocity,and angle.The system utilizes multiple transmitting and receiving antennas for multi-beam imaging,thereby improving the radar’s resolution and reliability.Compared to traditional radar systems,FMCW MIMO radar has higher degrees of freedom,higher measurement accuracy,and better antiinterference capability.Additionally,FMCW MIMO radar can simultaneously process multiple targets and image targets in different directions.Therefore,this takes advantage of FMCW MIMO radar and conducts research on parameter estimation algorithms based on FMCW MIMO radar.This thesis first introduces the theoretical basis of FMCW MIMO radar,including the FMCW modulation method and ranging principle,explains the impact of the number of elements on the phased array pattern,and derives the formation method of MIMO virtual arrays based on a uniform linear array.Then,it introduces the classical angle estimation method based on beamforming and analyzes its existing defects through simulation.Based on the characteristics of FMCW MIMO radar,this thesis constructs a mathematical model of FMCW MIMO radar signals and discusses parameter estimation methods based on FMCW MIMO radar.In response to the problem that the accuracy of frequency-based parameter estimation methods is limited,the feasibility of subspacebased super-resolution algorithms on FMCW MIMO radar is analyzed.Firstly,the Multiple Signal Classification(MUSIC)algorithm uses the orthogonality of the subspace and space-time conversion theory to construct spectral peak search functions for angle and distance independently,and estimates angle and distance parameters independently through grid traversal.For multi-target situations where one-dimensional subspace algorithms cannot perform parameter matching,the MUSIC algorithm is extended to two dimensions,using 2 Dimension Multiple Signal Classification(2D-MUSIC)for joint distance-angle estimation.This algorithm reconstructs the guiding vector,constructs a two-dimensional spectral function that contains both distance and angle information,and then selects a suitable two-dimensional search grid to search the two-dimensional spectral function to obtain pairs of distance-angle estimates.To address the problem of significant performance degradation of subspace-based algorithms in low signal-to-noise ratio and small pulse numbers,a multi-target localization algorithm for FMCW MIMO radar based on pseudo-noise resampling is proposed to improve the accuracy and stability of target localization in non-ideal environments.This algorithm first decomposes the covariance matrix,eliminates errors caused by non-ideal items through iteration,then forms a coarse estimation set by combining pairs of distance and angle estimates and performs threshold detection on it.The algorithm uses pseudo-noise resampling technology to eliminate outliers in the coarse estimation,achieving high-precision target localization under low signal-to-noise ratio and small pulse numbers.Finally,extensive experimental simulations are performed to verify the effectiveness and superiority of the proposed target localization algorithm.