DOA Estimation Algorithm And Application For MIMO Sonar

Multiple-Input Multiple-Output(MIMO)sonar is a new system,which employs multiple transmit sensors to emit orthogonal waveforms.The receive array adopts coherent processing for the echo signals,obtains more degrees of freedom and diversity gain.This paper starts with the reconstruction of covariance matrix,the signal processing of observations and the transmit signal energy optimal design,the DOA estimation technique for underwater targets based on MIMO sonar is studied.For the problem of noise interference,colored noise is simulated and analyzed,and a complex noise field closer to the actual marine environment is constructed.The noise is suppressed by methods such as dimension reduction and reconstruction of covariance matrix and beam space design.The proposed DOA estimation algorithms have the advantages of reducing computational complexity,improving direction-finding accuracy and enhancing robustness.It provides the theoretical basis and application foundation for the development of low-complexity MIMO sonar with stable direction-finding performance.The main research results and innovations of the paper are as follows:1.A signal model suitable for colocated MIMO sonar arrays is established,and the relationship between the direction-finding performance of MIMO array and system parameters is pointed out.By modeling and analyzing the transmit signals,virtual array elements and virtual aperture of the MIMO array,the corresponding beam pattern is drawn.The obtained results show that the MIMO array transmitting uncorrelated signals can obtain more effective virtual array elements and have better direction-finding performance.2.A SCAML optimization method based on MIMO-Capon algorithm is proposed.In order to avoid the increase of computational complexity caused by matched filtering processing,the feasibility of applying Capon algorithm directly to target parameter estimation is studied and analyzed,the signal model without matched filtering processing is established,and the MIMOCapon algorithm is obtained.For the situation that the MIMO-Capon algorithm has poor ability to estimate the target intensity,the maximum likelihood estimation method is used to modify the preprocessing results of the algorithm.The obtained SCAML optimization algorithm effectively makes up for the limitation that MIMO-Capon algorithm cannot improve the target strength estimation performance under the condition of high SNR.The related simulation experiments show that the MIMO-Capon algorithm effectively reduces the computational complexity,and the SCAML algorithm has higher direction-finding accuracy and target strength estimation accuracy.3.Aiming at the problem that the large-dimension covariance matrix reduces the algorithm’s direction-finding efficiency,fast MUSIC algorithm and fast ESPRIT algorithm are proposed.The signal model obtained by matched filtering processing can obtain additional degrees of freedom,but the computational complexity also increases.In order to increase the speed of the algorithm and ensure the number of detectable targets of the system,the MIMO array flow matrix is used to construct the dimension reduction transformation matrix,and the echo signals are weighted.The results of simulation and water tank experiments show that this method can effectively improve the direction-finding efficiency and performance of the algorithm.In the case of a limited number of snapshots,it has the characteristics of fast calculation speed and strong target resolution.4.A MIMO-STIM algorithm based on covariance matrix reconstruction is proposed,which can effectively suppress the impact of symmetric noise on the performance of the algorithm.First,a complex noise field model containing Gaussian white noise and color noise is established,and the noise field is divided into symmetric noise and asymmetric noise.According to the autocorrelation function characteristics of the symmetric noise component,a method to improve the performance of the algorithm by suppressing the noise component in the real part of the covariance matrix is obtained.Then,the real part of the covariance matrix is reconstructed by the dimension reduction transformation processing and matrix imaginary part replacement principle,which can avoid the interference of double spectrum.Finally,the Toeplitz method is used to coherently modify the reconstructed covariance matrix.The related theoretical analysis,simulation and water tank experiments show that this method can obviously suppress symmetric noise and achieve accurate direction-finding of weak signals.5.Aiming at the problem of complex noise suppression,an ME-STIM optimization method is proposed.Considering the influence of the accuracy prior knowledge and computational complexity on the performance of the algorithm,the ESPRIT algorithm is selected for pre-estimation calculation to avoid the angle search.Then,reconstruct the covariance matrix and use the fast MUSIC algorithm to perform the second DOA estimation,which can further suppress the asymmetric noise components.Corresponding simulation and water tank experiments prove that the ME-STIM algorithm has obvious ability to suppress complex noise.In the case of low signal noise,the direction-finding performance is superior to the MIMO-STIM algorithm.6.A data iteration algorithm based on LMS is proposed,which solves the situation that the decreasing direction-finding performance of covariance matrix algorithms in the case of limited snapshots,and avoids the computational complexity caused by the construction and inversion of the covariance matrix.This algorithm directly uses the snapshots to perform iterative calculations,searches and iterates the array flow pattern vector corresponding to each direction to complete the target direction finding.Simulation experiments and water tank experiments also show the superiority of the method that the algorithm is less sensitive to snapshots,more robust,and has better DOA estimation performance under the conditions of limited snapshots.Compared with traditional subspace algorithms,the LMS-based data iteration algorithm effectively reduces the computational complexity.7.A MIMO sonar beam design method is proposed,which effectively improves the direction-finding accuracy and target resolution of the algorithm.This method not only retains the diversity characteristics of MIMO sonar,but also obtains a portion of the phased array transmit coherent processing gain.By establishing constraint conditions to solve the optimal weights,the beam space is divided into sub-arrays to ensure the rotation invariance of the array and the uniform distribution of the transmission energy of each array element.Theoretical analysis and simulation results show that,compared with traditional subspace algorithms,the beam design method has higher direction-finding accuracy and target resolution ability in the case of low signal-to-noise ratio and limited number of snapshots,and has better ability to suppress Gaussian white noise and complex noise.The research results of the thesis effectively reduce the computational complexity of the system,and improve the direction-finding accuracy,target resolution and noise suppression ability of the algorithm under the premise of taking full advantage of the MIMO sonar array.Simulation and water tank experiments have verified the effectiveness of relevant technologies and results,these techniques can be applied to the engineering development of MIMO sonar.