Spatial Spectrum Estimation Algorithm For Coprime Array And Its FPGA Implementation

Spatial spectrum estimation,also known as Direction of Arrival(DOA),which is a key problem of array parameter estimation,is widely used in civil and military realms,such as communications sonars,radar.In array signal processing,uniform linear array,uniform circular array,L shaped array,uniform planar array are the common array structures.With the development of the theory and technology of array signal processing,DOA estimation for coprime array has become hot research recently.Compared with the traditional array,the spacing of coprime array can larger than half wavelength which can get larger array aperture with the same number of elements.It also has a higher degrees of freedom,so that it will have a higher direction accuracy and resolution.Therefore,the research on spatial spectrum estimation for coprime array has a high theoretical value and application prospect.The main work of this paper is summarized as follows:(1)Three one-dimensional DOA estimation algorithms for coprime linear array are investigated: Multiple Signal Classification(MUSIC)algorithm,Estimating Signal Parameters via Rotational Invariance Techniques(ESPRIT)algorithm,Propogator Method(PM).Under the condition of equal number of array elements,the three algorithms have better performance than the same algorithm for uniform linear array and with lower complexity.(2)Three 2D-DOA estimation algorithms for coprime planar array are investigated: 2D-MUSIC algorithm,2D-ESPRIT algorithm,2D-PM algorithm.The 2D-MUSIC algorithm has high complexity for total spectral search.ESPRIT algorithm don't need spectral search and achieves paired elevation and azimuth automatically.PM algorithm does not require spectral peak searching and eigenvalue decomposition of the covariance matrix of received signal,so it has low complexity.With equal number of array elements,the three algorithms which are investigated have better performance than the same algorithm for uniform planar array and with lower complexity.(3)Three 2D-DOA estimation algorithms for coprime L-shaped array are investigated: 2D-MUSIC algorithm,2D-ESPRIT algorithm,2D-PM algorithm.Although the 2D-MUSIC algorithm has the merits of good estimate precision but global spectral searching brings high complexity.2D-ESPRIT algorithm without spectral peak searching and they can also achieve automatically paired 2D-DOA estimation.2D-PM algorithm does not need spectral peak search while avoiding the eigenvalue decomposition of covariance matrix,has better two-dimensional DOA estimation performance and low complexity.The three algorithms which are investigated have better performance and lower complexity than the common algorithm for uniform planar array with the same number of array elements.(4)Realized one-dimensional DOA estimation based on MUSIC algorithm and ESPRIT algorithm for coprime linear array on PFGA.Including FPGA design for covariance matrix construction,eigenvalue decomposition,matrix inversion,spectral peak search,ambiguous DOAs elimination modules.The Modelsim simulation results of above modules are given and shows hardware test results of MUSIC algorithm and ESPRIT algorithm.Simulation and test results verify the effectiveness and correctness of the FPGA design.