| In recent years,the application of millimeter wave radar has expanded across various fields,including unmanned driving,traffic monitoring,target detection,and positioning.This technology offers several advantages,such as a wide detection range and all-weather capability,remaining unaffected by factors like temperature,illumination,rain,and snow.In target detection and localization,accurate estimation of the direction of arrival(DOA)is a crucial step.Existing DOA estimation methods,such as Fast Fourier Transform(FFT),Multiple Signal Classification(MUSIC),and other subspace algorithms,have been employed to achieve DOA estimation.However,the FFT algorithm often exhibits poor estimation accuracy,while the MUSIC algorithm suffers from high computational complexity,which can impact the Angle measurement performance of millimeter wave radar.To address the issues of poor estimation accuracy and high computational complexity in existing DOA estimation methods,this thesis focuses on developing an improved algorithm that offers high accuracy while reducing computational requirements.The goal is to enhance the accuracy of millimeter wave radar in Angle measurement while minimizing computational overhead.The research work carried out in this thesis is as follows:Firstly,the thesis introduces the transmission,echo,and beat signals of millimeter wave radar using Modulation Continuous Wave(FMCW)signal as an example.Based on spatial spectrum estimation,the DOA estimation principles of traditional FFT algorithm,MUSIC algorithm,and ESPRIT algorithm are discussed.Furthermore,the operational complexity of these algorithms is analyzed,and the impact of different factors on the algorithm is studied through simulation.Then,in order to address the issue of poor accuracy in DOA estimation using the FFT algorithm,this thesis extends the Rife algorithm from time domain frequency estimation to spatial DOA estimation.Through simulations,it is demonstrated that the accuracy of FFT-based Angle measurement can be effectively improved.However,the Rife algorithm lacks super resolution,and there are still errors in Angle measurement accuracy.To overcome these limitations,the thesis proposes the Rife-MUSIC algorithm.The idea is to use the Rife algorithm to divide the Angle estimation range and restrict the MUSIC algorithm to search within a smaller range,significantly reducing the computational complexity associated with the traditional MUSIC algorithm.Theoretical analysis and extensive simulations show that the performance of the Rife-MUSIC algorithm surpasses that of the classical MUSIC algorithm,particularly in scenarios with low signal-to-noise ratio.Moreover,the Rife-MUSIC algorithm preserves the super resolution capability of the MUSIC algorithm,making it highly promising for practical applications.Recognizing that the key aspect of the MUSIC algorithm is the search for spectral peaks corresponding to Angle values,an improved approach is proposed that utilizes a quadratic interpolation method to locate the extreme values associated with Angles based on the estimated Angle range obtained from FFT preprocessing of the noise subspace.This method ensures accuracy and eliminates the need for tedious step scanning calculations.Theoretical analysis confirms that the improved algorithm significantly reduces operational complexity.The proposed algorithm’s effectiveness is verified through simulations,and the impact of various factors is analyzed.Finally,field tests are conducted to validate the proposed algorithm using an actual millimeter wave radar platform for data collection.These tests provide practical evidence of the algorithm’s effectiveness and its real-world applicability. |
