Research On Direction Finding Algorithm Of Instantaneous Signal

With the increasing complexity of the electromagnetic environment in space and the demand for fast and real-time target perception,the instantaneous signal direction finding scenes under extreme single snapshots are becoming more and more extensive.In military communications,due to the unknown hopping mode or short duration of frequencyhopping signals and burst signals which designed for anti-reconnaissance,the array can only intercept the signal data with single snapshots.In addition,highly maneuverable platforms such as airborne and shipborne have very high requirements for the real-time performance of the direction finding system,and it is impossible to collect snapshot data with long time sampling.The traditional DOA estimation method cannot be directly applied to instantaneous signal direction finding because the data is not enough to accurately divide the subspace with single snapshot,most of the existing single snapshot estimation methods are based on one-dimensional linear array,which can only estimate one-dimensional angles.Through the study of rank restoration and two-dimensional DOA estimation,this paper studies the two-dimensional DOA estimation method with single snapshot based on parallel linear array and L-shaped array.The research work of this paper is as follows:For instantaneous signal direction finding with dual parallel linear array,the two rank restoration methods of forward-backward spatial smoothing(FBSS)and constructing pseudo-covariance matrix on one-dimensional linear array are respectively extended to dual parallel linear array,derive the reconstruction process of single snapshot data and analysis the form of the equivalent covariance matrix,combined with the 2-D ESPRIT method,two single snapshot direction finding algorithms with dual parallel line array based on FBSS pre-processing and constructing pseudo-covariance matrix are presented.The computational complexity of these two dual-parallel linear array singleshot direction finding algorithms is analyzed,and compared with the SS-DOAM algorithm in other documents in terms of target resolution and statistical performance.A single snapshot direction finding algorithm based on constructing pseudo-matrix with three parallel linear array is studied to reduce the loss of freedom of the single snashot algorithm based on dual parallel linear array.The construction method of pseudo covariance matrix is modified,three full rank pseudo matrices are constructed based on three parallel linear matrices,and the two-dimensional DOA is solved based on the rotation invariance between pseudo matrices.Under the same total number of array elements,the degree of freedom and statistical performance of the three-parallel linear array algorithm are better than those of the dual parallel linear array single-shot direction finding algorithm.For instantaneous signal direction finding with L-shaped array,based on the study of the conventional L-shaped array ESPRIT,a new method of ESPRIT for L-shaped array is studied by designing a new large matrix construction form for subspace decomposition.Compared with the conventional method,this method removes the noise items existing in the large matrix while maintaining a lower matrix dimension,and achieves a better balance between the calculation cost and the estimation performance.By extending the FBSS pre-processing to the L-shaped array and combining with the new L-shaped array ESPRIT,a single-shot direction finding algorithm of L-shaped array based on the FBSS pre-processing is studied.A single snapshot direction finding algorithm of L-shaped array based on constructing pseudo-matrix is studied.The pseudo-matrix construction form of three parallel linear array is extended to L-shaped array by constructing four full-rank pseudomatrices on L-shaped array,and use the rotation invariance between the matrices to estimate the two-dimensional angle.The algorithm only has one degree of freedom loss caused by ESPRIT,and the target resolution and statistical performance are better than the single snapshot direction finding algorithm based on FBSS pre-processing.