Research On High Resolution Imaging And Interference Suppression For Forward Looking Sonar

High frequency forward looking sonar can visually present the underwater target images,which has been widely used in the fields of underwater small target detection,auxiliary navigation and underwater search rescue.Interference suppression and high resolution imaging are two key technologies for forward looking sonar.First of all,the interference may cause the background elevation of the sonar images,resulting in the decrease of the dynamic range of the sonar image,affecting the sonar imaging quality,and reducing the detection performance of the small and weak targets.Secondly,when the sonar sidelobe is higher,the adjacent weak targets are submerged in high sidelobe caused by the strong interference and are missed detection.Thirdly,the false targets are produced caused by the strong interference,so that it is false alarm.Multiple input multiple output(MIMO)imaging sonar is an effective way to achieve high resolution,which has been attented extensively by the scholars at home and abroad.However,the superposed echoes for the multiple transmitted orthogonal waveforms cannot be separated perfectly for conventional MIMO imaging sonar,which may severely reduce the quality of underwater acoustic images because of cross‐correlation noise.In this paper,the interference suppression and crosstalk suppression of the superposed echoes for the multiple transmitted orthogonal waveforms are studied to improve the performance of forward looking sonar imaging.The main research contents include the following aspects:The research on the imaging method of two-dimensional low sidelobe based on spatially variant apodization for forward looking sonar.Sidelobe suppression using conventional window function weighted is invariably at the expense of mainlobe resolution,in order to solve this problem,and the sonar imaging method based on improved spatially variant apodization is poposed.In this method,spatially variant apodization method is applied to the post-processing of the azimuth-range two dimensional beam domain data,and also the magnitude and phase errors calibration combined with spatially variant apodization is to avoid the sidelobe elevated.So the robustness of the spatially variant apodization method is improved,and the sonar imaging effect of two-dimensional low sidelobe level and unchanged resolution is achieved.The results of numerical simulations and a lake experiment demonstrate that the quality of the sonar images are greatly improved,which the sidelobes are effectively suppressed of azimuth-range dimensions and also the resolution is not sacrificed.The research on the interference suppression method in sonar imaging utilizing correlated projection and eigenspace processing.The capability of orthogonal projection approach degrades severely in presence of array model mismatch,especially when the training samples are mixed with the desired signal.Therefore,an improved orthogonal projection robust adaptive beamforming is proposed based on correlated projection and eigen-space processing.In the proposed approach,the interference subspace is constructed combining the correlated projection and eigen-space processing firstly.Then the interference-plus-noise covariance matrix is accurately reconstructed to eliminate desire signal from sample covariance matrix.Subsequently,the desired signal steering vector is corrected applying correlated projection and then the adaptive weighted vector is modified by orthogonal projection approach.Finally,the modified adaptive weighted vector is applied to the forward looking sonar imaging process.The results of numerical simulations and a lake experiment demonstrate when the training samples of imaging sonar contain both interferences and desired target signal,the proposed method can not only suppress the interference effectively,but also retain the desired target signal.As a result,the robustness of the orthogonal projection adaptive beamforming is improved.The research on suppression method of the cross‐correlation noise between the superposed echoes in MIMO imaging sonar based on fractional Fourier transform.In practical,it is difficult to obtain completely orthogonal waveforms for MIMO imaging sonar and the cross‐correlation between different orthogonal waveforms is not equal to zero,which may lead to ranging ambiguity and reduce the quality of underwater acoustic images.To solve this problem,an orthogonal waveform separation method based on the fractional Fourier filtering to suppress the cross‐correlation noise between received orthogonal waveforms is proposed for the MIMO imaging sonar.First,the superposed echoes for transmitted linear frequency‐coded waveforms are processed using a fractional Fourier transform.Then,time‐varying fractional Fourier filters are used to separate echoes corresponding to different transmitted waveforms.Subsequently,images are obtained by joint transmit‐receive beamforming for the output of matched filtering with weighting and ranging ambiguity is avoided.Finally,the results of numerical simulations and a lake experiment demonstrate that the cross‐correlation noise between the received orthogonal waveforms is effectively suppressed.Moreover,the image becomes sharp and its quality is improved with sidelobe levels lower than-40 d B.The research on beam sidelobes suppression method in MIMO high resolution imaging sonar.For MIMO imaging sonar,a spanned virtual array with a larger aperture is formed to enhance the spatial resolution and image quality.However,the conventional window function weighted method to reduce beam sidelobe levels is not suitable in MIMO imaging sonar due to the non-uniform of the virtual array.To solve this problem,the low beam sidelobe method for MIMO imaging sonar based on window function coefficient corrected is proposed.In the proposed method,the constraints of desired target signal intensity and beampattern response sensitivity is introduced,and the second-order cone programming constraint optimization problem is solved.Then,the corrected window function coefficient is applied to the MIMO sonar imaging process to achieve beam sidelobe suppression.Numerical simulations and a lake experiment are presented to demonstrate that the proposed method can effectively suppress the beam sidelobe for MIMO imaging sonar.Furthermore,the MIMO sonar imaging method is proposed based on beam domain deconvolution.The relationship between the MIMO sonar echo signal model and the beam response is deduced,and then the beam domain deconvolution method with the acceleration iteration of the predicted direction vector is used to obtain the low sidelobe beam response.Numerical simulations and a lake experiment are presented to demonstrate that the proposed method can effectively reduce the sidelobe levels lower than-40 d B and simultaneously decrease the mainlobe width.The research work in this paper can effectively enhance the performance of forward looking sonar imaging and also improve the quality of underwater acoustic images,which is conducive to the subsequent underwater dim small target detection and recognition.