Acoustic Signal Modeling And Processing For Optic-acoustic Sea Bottom Detection

Ocean remote sensing is an ocean observation method which uses non-contact sensing techniques. It is used to acquire image or data of ocean landscape, components, and/or processes to monitor the ocean environment, detect ocean bottom and observe dynamic phenomena. Currently widely used bathymetry measurement systems include the single-beam echo sounder, multi-beam echo sounder and airborne laser sounder. Those devices have their own merits as well limitations respectively.The concept of optic-acoustic detection breaks the limitations of both traditional optical depth measurement and acoustic depth measurement. To generate laser sound pulse as a detection signal, a high energy laser pulse is used to hit the water surface. The fiber hydrophone based on acousto-optic coupling is exploited as the echo signal receiving device. This method integrates the sonar system’s resolution and detection range advantages with speed and mobility of an airborne platform. In numerous applications, it can be an important supplement to traditional topographical measurements.This thesis studies the acoustic signal modeling and processing in optic-acoustic sea bottom detection. Optic-acoustic sea bottom detection uses new detection signal and new method of signal receiving, which bring convenience as well new problems. In this thesis, we focus on the seabed backscattering signal modeling of the laser acoustic pulse, estimation of the time-of-arrival (TOA) of the echo signal, and suppression of the interference caused by the waves.In active sonar system, the characteristic of the transmitted signal is a vital part. This thesis studies the characteristics of the laser acoustic pulse signals by analyzing the experimental laser acoustic pulse data. On that basis, we develop a seabed backscattering signal model of the laser acoustic pulse. The model integrates the characteristics of the laser acoustic pulse with propagation attenuation, seabed scattering function and a few other factors. The rationality of the model is verified by simulations.Bottom detection is a hypothesis testing problem with unknown parameters. The processing signal is the seabed backscatter echo signal. This thesis uses the characteristic function correlation detection to estimate the TOA of the echo signal. The characteristic function in the method is the envelop sequences of the simulated seabed backscattering signal. The characteristic parameter correlation detection method utilizes the features of the amplitude and envelope backscattering signal. The simulation results show the effectiveness and the anti-interference capability of the method.Finally from the perspective of signal receiving, this thesis studies underwater acoustic signals from an acoustic-optic detection experiment using acousto-optic coupling principle, particularly analyzing the interference caused by the waves. In the framework of Bayesian signal processing, a time-varying AR model based Kalman filter is designed to track and suppress the wave interference signal. By examing the experimental data filtering results, the performance improvment of the designed adaptive filter over the traditional bandpass filtering method is verified.