Direction Of Arrival Estimation Based On Co-prime Array

Direction of arrival(DOA)estimation is now a research hotspot.At present,most of DOA estimation algorithms are based on a uniform array with equal array element spacing,and many classical algorithms are proposed based on the receiving signal model of uniform array.However,the interval between two elements of uniform array structure is limited to half wavelength,and once the interval exceeds the limit,the estimated result will be angular ambiguity.Such spacing limitation has two effects on DOA estimation performance: one is that the array aperture is small due to the limited array spacing,which affects the estimation accuracy,the other is that the distance between arrays is too small,resulting in the mutual coupling effect between arrays,which affects the estimation results.In response to these shortcomings,many researchers consider changing the structure of arrays and begin to study heterogeneous arrays.The spacing of non-uniform arrays is no longer limited so the aperture of the array is enlarged and the mutual coupling effect is reduced.There are many structural forms of heterogeneous arrays,among which the mutual-quality array is one.The work of this paper is to carry out a series of researches on the DOA estimation of the mutual-quality array:In the first part,a decoherence algorithm is proposed for DOA estimation of coherent signals.The main idea is to use the electromagnetic vector sensor as the array element of the coprime array,and solve the rank loss problem of coherent signals based on multiple receiving components of the electromagnetic vector sensor.On the other hand,the virtual array element derived from the virtual domain of the coprime array is not continuous,which will cause information loss,therefore,the algorithm also combines the idea of matrix element interpolation to optimize covariance matrix,and the data is recovered to form a continuous virtual array.Compared with the original virtual array,its aperture is expanded,so the performance of the algorithm is improved.In the second part,two kinds of array are used to estimate the two-dimensional incident signals.One is the improved L-shaped array.The improved L-shaped array is used to construct an L-shaped antenna array to realize the two-dimensional angle estimation.The other is the parallel coprime array.By constructing two correlation matrices and processing them,the received data of a group of virtual array elementscorresponding to the actual array element position and difference set are obtained.The number of virtual array elements is increased,the aperture of virtual array is expanded,and the degree of freedom and accuracy of DOA estimation are improved.In the third part,multi parameter estimation of near-field source based on the coprime array is proposed for the near-field incident signal.Nowadays,most of the near-field source positioning only estimates its elevation angle and distance.In this algorithm,the angle estimation of the near-field incident signal is extended to three-dimensional space,and the elevation angle,azimuth angle and distance of the near-field incident signal are estimated by using the coprime array to receive the signal.The special array structure is used to transform the three-dimensional parameter estimation problem into three one-dimensional parameter estimation problems,and sparse reconstruction algorithm is used to estimate.The algorithm itself has good noise robustness,and does not need to use spatial smoothing to cause aperture loss,so it can achieve better estimation performance.