Performance Improvement Of Source Number Estimation Based On UCA And Its FPGA Implementation

In the array signal processing,the estimation of the number of signal sources is an important part of spatial spectrum estimation.The direction of arrival(DOA)estimation performance will deteriorate rapidly when the number of estimated sources is not consistent with the number of real sources.Therefore,accurate estimation of the number of sources is a key techniques for DOA estimation.The estimation performance of the existing algorithms will decline when the number of sources is close and even more than the number of elements in the colored noise environment.Under the condition of UCA,an algorithm for estimating the number of signal source based on virtual array is proposed.The main contents of this paper include:1.Taking the MUSIC algorithm as an example,the influence of error estimation of source on DOA is studied.A brief description of the conventional source numb er estimation algorithm is given,and the influence of source number estimation is verified under the change of the array element,color noise correlation,SNR and snapshot number.2.According to the number of sources close to the number of elements,there are many algorithms to estimate performance degradation due to the lack of freedom.It's possible to increase the number of detectable signal sources by virtual array element.In this paper,the interpolation transform and the fourth-order cumulant of the virtual antenna technology are studied,it is verified that the virtual antenna increases the freedom of the array.The fourth-order cumulant matrix is constructed for minimum redundancy in UCA,and the influence of array expansion on the estimated performance of the number of source is verified by combining Gerschgorin disks estimation(GDE)method under white noise.3.Due to the uncertainty of the color noise environment,and the fourth-order cumulant matrix defects under limited snapshot,there is no obvious difference between the signal Gail disk and the noise of the Gail disk,which makes the adjustment factor sensitive and affects the estimation performance.The Gail disk criterion and the fourth-order cumulant can overcome the influence of color noise.In this paper,an improved algorithm of Gail based on fourth-order cumulant matrix of minimum redundancy is proposed.Avoid the impact of human factors,so the method of automatic adjustment factor is designed.Finally,the complexity of the algorithm is evaluated,and the effectiveness of the proposed algorithm is verified by simulation and experimental data.4.The FPGA of Kintex-7 series implementation of the source number estimation algorit hm is discussed.The FPGA implementation of each module is analyzed.Finally,the system resource evaluation is carried out,and the FPGA test results are analyzed in the outfield environment experiment.The method presented in this paper is embodied in two aspects.On one hand,the virtual array generated by the four order cumulant can make up for the deficiency of the system freedom when the number of sources is close to the number of elements in Gail disk.On the other hand,the problem of the divergence of the noise eigenvalue in the four order cumulant matrix can be overcome by the improvement of the Gail radius compression...