Research On DOA Estimation For Random Sub-nyquist Sampling In Array Signal Processing Based On Matrix Completion

DOA estimation which is an important branch of array signal processing has been widely used in electronic surveillance, radar jammers and radar positioning. At the present stage, although the performance of DOA estimation can be improved by increasing the number of the array antennas, elevating the rate of sampling and using wideband or ultra-wideband signal, at the same time, the pressure of sampling, transmission and storage of array signal processing system will increase, which may enhances the complexity of the system. This thesis mainly research how to mitigate the problem of the system complexity without changing the performance of DOA estimation. The author's major contributions are outlined as follows:Firstly, this thesis introduces the basic knowledge, the common structure and the receiving signal model of the array signal processing, especially introduces the receiving signal model based on linear array in detail. The algorithm include MUSIC and ESPRIT that based on subspace decomposition algorithm is used to DOA estimation, and it's principle and performance analysis are given in the thesis detailedly. The feasibility of the application of Matrix Completion in array signal processing is been studied, and algorithms of SVT, ALM and APG to solve the problem of Matrix Completion are been analyzed and simulated in MATLAB.Secondly, under the background of DOA estimation in array signal processing, in consideration of the problem that exists in the current research on DOA estimation, this thesis apply the matrix completion to DOA estimation to slove the problem of system complexity.Finally, the principle and the implementation step of the method of down-sampling DOA estimation based on snapshot are listed in this thesis. The feasibility and the performance of DOA estimation of this algorithm are proved in using theory simulation. The method of down-sampling DOA estimation based on Hankel is presented, which the Hankel matrix is been applied in DOA estimation for the first time in this thesis. This method applies the matrix completion to complete the missing data, with the low rank property of the Hankel matrix which is constructed by the each row of the down-sampling matrix of array signal, the missing data can be completed in high probability, and the performance of DOA estimation can be gained. This algorithm adoptes the method that decline the rate of sampling to decrease the complexity of array signal processing system. Simulation results show the effective of the proposed method which can improve the performance of DOA estimation, and can decrease the complexity of array signal processing system.