Research On Target Orientation Algorithm Based On Sensor Array Extension

The direction of arrival(DOA)estimation is a very important research element in array signal processing.In recent years,the DOA estimation techniques have developed rapidly and can achieve high-resolution angular estimation of multiple sources in different directions in space.With the development of related research and technology,the performance of DOA estimation algorithms has become more and more demanding,and some existing high-precision spatial spectrum estimation algorithms are usually computationally intensive.In this thesis,we consider the research around reducing the computational complexity of the algorithm and enhancing the real-time and estimation performance of the algorithm in the case of low signal-to-noise ratio.The main contributions and innovations are as follows.First,the current research status of the DOA estimation algorithms is analyzed.The principle of DOA estimation based on the MUSIC algorithm,the ESPRIT algorithm and the OPM algorithm are introduced.,and their operational complexity is analyzed through simulation experiments.For the problem that the MUSIC algorithm cannot accurately distinguish the coherent signal,the spatial smoothing algorithm and the Toplitz approximation algorithm are introduced.The experiments show that both algorithms can achieve a good decoherence effect and overcome the drawbacks of the MUSIC algorithm.Secondly,the fourth-order cumulative MUSIC algorithm(MUSIC-LIKE),the fourth-order cumulative ESPRIT algorithm and the fourth-order cumulative OPM algorithm(FOC-OPM algorithm)are analyzed from the perspective of expanding the array aperture and improving the array element utilization.The simulation results show that these three fourth-order accumulation algorithms effectively extend the array aperture and make full use of the received array element data.In addition,the adaptability of the algorithm to conditions with low signal-to-noise ratio is effectively improved due to the suppression of Gaussian noise by using fourth-order accumulation.Finally,in order to improve the estimation accuracy and resolution,two target vectoring algorithms combining the fourth-order accumulation of virtual array elements,namely,the improved fourth-order MUSIC algorithm and the improved fourth-order OPM algorithm,are proposed in this thesis.The basic principle of the improved algorithm is to use virtual a reference array element,on the basis of reconstructing the new guiding vector matrix and the fourth-order cumulants matrix,removing the data redundant information of the fourth-order cumulants in the uniform equidistant line array estimation,and combining with the traditional MUSIC algorithm and the OPM algorithm respectively,so as to achieve the purpose of array target orientation.The new array expansion method of virtual array elements improves the utilization rate of the received data of the array elements.The experiments show that the two improved fourth-order cumulants algorithms proposed in this thesis achieve the expected goal of improving the DOA estimation accuracy and reducing the operational complexity of the traditional fourth-order cumulants algorithm while realizing the array expansion.