Research On Acoustic Vector Sensor Array For Calibration And Direction Finding Techniques

Compared to scalar array,acoustic vector sensor array(AVSA)overcome the starboard ambiguity,expand array aperture and reduce background interference and detection threshold,it have obvious advantages in the case of online array,planar array or few array elements.Based on the above,AVSA signal processing techniques become a hot spot in underwater acoustic field,United States and Russia vigorously exploit AVSA to serve naval equipments,which implement the detection and localization for underwater targets,the importance of modern acoustic vector array of naval equipment is evident.This paper researched the calibration and DOA estimation technique based on AVSA.High-resolution performance of numerous DOA estimation algorithms based on AVSA was obtained by the ideal array manifold,in practical,there is a variety of array errors accompany-ing AVSA,which will cause the performance of amount high-resolution algorithms seriously degenerate and even effectiveness.Therefore,calibration for AVSA is essential before using the high-resolution DOA estimation algorithms.For direction finding problem by AVSA,most DOA estimation algorithms based on AVSA have a large amount of computation,which greatly increase the cost and burden of systems,even become a bottleneck to some systems.In addition,these are many coherent signals in underwater acoustic environment,researching muti-target resolution of coherent sources is significance for AVSA direction finding techniques.The main contents of this paper are:1.Consider calibration for gain,phase and location errors,a new fast algorithm named SMSWF based on MSWF was proposed.SMSWF algorithm take advantage of the DOA and waveform of the calibration source to estimate the gain,phase and location factors without estimating the covariance matrix and eigendecomposition,compared with eigendecomposition algorithm,SMSWF algorithm has the same estimation performance of gain and phase factor while greatly reduce the computational complexity.The research showed that if a single source incident on the array and its waveform is known,signal subspace obtained by SMSWF and eigendecomposition are equipollence,which demonstrate that SMSWF could replace eigendecomposition and the computational complexity of many signal processing methods which contain eigendecomposition would greatly reduced by replacing eigendecomposition with SMSWF.Extensive computer simulations and experiment result in anechoice water tank shows the superiority performance of SMSWF algorithm.2.Consider the attitude errors of acoustic vector sensor array elements and its calibration,theoretical analysis and simulation analysis about the influence of array elements attitude errors acting on the beam pattern and MUSIC algorithm based on AVSA was respectively presented.In the actual work environment,the calibration algorithm with known sources was always restricted,a self-calibration algorithm of array elements attitude errors was proposed,this algorithm can estimation the attitude errors parameters and DOA of sources Simultaneously,which has good performance of parameters estimation and fast convergence rate.Theoretical analysis and computer simulations verified the superiority of the self-calibration algorithm.3.Consider the problem that huge computation of high-resolution DOA estimation algorithm using acoustic vector sensor array,V-MSWF and PV-MSWF algorithms was proposed.V-MSWF is a spread form scalar array to AVSA.PV-MSWF is based on the combination processing of pressure and particle velocity,which selected the electronic rotation vector of the reference element as the desired signal and MSWF is used to estimation the signal subspace,PV-MSWF is based on the principle of coherency between pressure and particle velocity,which can suppress interference well in isotropic noise field.V-MSWF and PV-MSWF algorithms both reduce the amount of computation because they did not need to calculate the cross-covariance matrix of acoustic vector sensor array and eigenvalue decomposition.Computer simulations and experiment result in anechoice water tank showed that V-MSWF and PV-MSWF algorithms have good performance of DOA estimation.4.PVFS(Particle Velocity Field Smoothing)algorithm based on acoustic vector sensor array is an effective decorrelation algorithm,but when there are a large number of coherent sources,its performance would degrade seriously and even fails.Based on PVFS,the MSS(Matrix Square Smoothing)-PVFS algorithm is proposed as the amelioration of PVFS.The proposed algorithm squares and partitions the data covariance matrix constructed by PVFS,then the partitioned matrix is cross multiplied each other.Finally,the decorrelation ability of PVFS algorithm is enhanced and more coherent sources can be distinguished.The computer simulations and experiment result in anechoice water tank indicates the superior DOA estimation performance of the proposed algorithm for coherent sources.