Application Of Compressed Sensing In Ocean Acoustic Tomography

Ocean acoustic tomography is based on the fact that the underwater acoustic signal will be affected by factors such as the temperature and velocity of seawater when it propagates in the ocean.It uses the speed and time of acoustic signal propagation in the ocean to invert the relevant parameters of the surveying sea area.This method is not limited by shipping,weather,and fishing activities.It can realize large-scale,long-term,continuous real-time monitoring of the ocean.However,there are still problems in the discrimination of different ray of acoustic propagation multipath,and the identification of corresponding ray propagation time.For example,the classical ocean acoustic tomography system uses a matched filtering method to obtain an estimate of the maximum signal-to-noise ratio acoustic signal arrival time,but the resolution of the recognition time structure is limited by the bandwidth of the underwater acoustic channel,and it is difficult to identify the multipath signal with smaller delay interval.This paper proposes the compressed sensing for time extraction,as a new signal processing method,it can use the sparse characteristics of underwater acoustic propagation channels to improve the estimation performance of multipath structures.Combined with the traditional matched filtering method,based on the sparse characteristics of the underwater acoustic propagation channel and the compressed sensing theory,the matched filtering-compressed sensing combination method is proposed to estimate the propagation delay time and amplitude information of the marine underwater acoustic signal,and to improve detection signal to noise ratio and multipath delay resolution for underwater acoustic signals.The simulations of underwater acoustic signal propagation were performed by Bellhop model.The performance of matched filtering method,direct compressed sensing method and matched filtering-compressed sensing combination method were evaluated respectively.It was confirmed that the matched filtering-compressed sensing combination method has theoretical advantages in extracting the arrival time of the multipath propagation of the underwater acoustic signal and the signal-to-noise ratio of the signal detection.Through processing and analysis of measured data,it further confirms the effectiveness and high accuracy of the matched filtering-compressed sensing combination method in improving the signal arrival time resolution and acoustic signal detection signal-to-noise ratio,so as to provide powerful data support for high-accuracy ocean acoustic signal propagation and acoustic tomography inversion of the temperature and flow field.