Research And Application Of DOA Estimation Algorithm In Vehicle Positioning

Internet of vehicle(IOV)is an important foothold of intelligent transportation system(ITS)in real-world applications.High-precision vehicle positioning,as one of the key technologies in IOV,has gradually aroused people’s research interest and became a current research hotspot.Direction of arrival(DOA)estimation technology is widely used in radar systems,sonar,wireless communications,medical imaging and other fields due to its high-precision positioning performance.With the explosive growth of people’s demand for IOV and ITS,DOA estimation technology is applied to vehicle positioning due to its high-precision positioning performance.Due to the multipath transmission effect in the actual application scenarios of vehicle positioning,the source correlation leads to a sharp decline in the performance of DOA estimation,which seriously reduces the accuracy of vehicle positioning.When the number of targets is greater than the number of antennas in the receiving array during vehicle positioning,the algorithm will not be able to estimate the DOA of all targets,resulting in missed detections,and will seriously reduce the accuracy of vehicle positioning.In addition,due to the real-time requirements of IOV for vehicle location information,it has higher requirements on the computational complexity of the DOA estimation algorithm.Therefore,this paper focuses on the problems in vehicle positioning based on DOA estimation in the IOV scene,and studies the decorrelation algorithm and low computational complexity algorithm of DOA estimation.The detailed research arrangements are as follows:(1)Analyze the vehicle positioning model and actual application scenarios.Aiming at the signal correlation in the IOV and the large number of estimated targets,a DOA estimation algorithm based on the fourth-order cumulant is proposed.The algorithm first uses the fourthorder cumulant and the idea of array division to construct a Toeplitz matrix.After eigenvalue decomposition,the eigenvectors of this Toeplitz matrix are orthogonal to each other,which effectively realizes the decorrelation of related information sources.In addition,the algorithm uses the blind Gaussian property of the fourth-order cumulant and the extended array aperture characteristics to effectively suppress Gaussian white noise and color noise,improve the performance of DOA estimation,and solve the problem that when the number of target positioning in the vehicle positioning system is greater than the number of antennas in the receiving array,the DOA of all targets cannot be estimated.Finally,the simulation experiment verifies that the DOA estimation algorithm of the correlation source based on the fourth-order cumulant proposed in this paper can not only realize the decorrelation of the source,but also has higher DOA estimation accuracy.However,because the algorithm uses a fourth-order cumulant,the computational complexity of the algorithm increases.(2)Aiming at the problem that the proposed DOA estimation algorithm based on the fourth-order cumulant for related sources has high computational complexity,which is not conducive to real-time vehicle positioning.On the basis of the proposed algorithm,this paper proposes a low computational complexity DOA estimation algorithm for related sources by improving the steps with higher computational complexity in the original algorithm.The improved algorithm and the positioning algorithm are combined to obtain a vehicle positioning algorithm with low computational complexity.Specifically,the vehicle positioning algorithm uses linear operations instead of eigenvalue decomposition,which effectively reduces the computational complexity of the algorithm.In addition,the algorithm uses the spatial spectrum function to be a breakpoint at the DOA,and the physical properties approach a larger value.A new "spatial spectrum" is constructed by calculating the first derivative of the spatial spectrum function for DOA estimation,which effectively improves the estimation accuracy of DOA under low SNR.Finally,the algorithm is combined with the vehicle positioning algorithm to get the location information of the vehicle.Simulation experiments and computational complexity analysis show that the low computational complexity vehicle positioning algorithm proposed in this paper not only effectively reduces the computational complexity of the algorithm,but also improves the accuracy of vehicle positioning under low SNR.