Hydroacoustic signal processing

The main task of signal processing is to sample the water sound field in time and space under the background interference, and transform it in space and time to improve the ability of detecting the required signal. In the early 1950s, with the development of information theory, signal detection theory and computing technology, the technology and theory of underwater acoustic signal processing also developed rapidly. By the early 1960s, underwater acoustic signal processing had mastered many technologies such as spectrum analysis, correlation, matched filtering and multi-beam forming.

With the rapid development of computer, underwater acoustic signal processing has made the following progress:

① The successful application of digital technology and adaptive control in beamforming has brought the spatial processing into a new stage-the multi-beam reception is becoming more and more perfect and can be adaptively matched with the environmental interference field. From simple space processing to optimal space-time processing.

② Considering the random characteristics of underwater acoustic transmission channel changing with time and space, which causes dispersion in time and frequency, we have been working hard to solve the problem of receiving and channel matching.

(3) The target recognition made a breakthrough and began to enter the practical stage. (4) Computer-aided target detection and tracking and parameter estimation have appeared, and they are developing towards automatic detection.

Spatio-temporal optimal processing is to maximize the use of spatial and temporal information, that is, the difference of spatial distribution of signals and background noise and the difference of time or spectral characteristics:

In terms of space, the conventional processing only uses the information of signals from a certain direction (through time delay matching and in-phase superposition), while the optimal spatio-temporal processing also uses the spatial correlation characteristics of noise fields. Compared with conventional processing, it has a spatial pre-white filter, whose function is to realize "noise cancellation" by using noise correlation, so as to eliminate the spatial correlation among various noises as much as possible and weaken the noise power.

As far as time is concerned, conventional processing only uses the spectral level difference of signal and noise power spectra; In addition, the optimal processing also takes advantage of the shape difference of signal and noise power spectra. According to the shape of signal and noise power spectrum, the optimal spatio-temporal processor constitutes the optimal preselection filter.

It can be seen that when the interference field is correlated interference, the optimal spatio-temporal processor will show significant advantages over the traditional spatio-temporal processor. In the case of resisting plane wave interference, the structure of the optimal spatio-temporal processor can be divided into three parts:

In the first part, the plane wave interference is eliminated. Firstly, the plane wave interference is estimated by matrix filter, and the estimated value of the space-time sample of the plane wave interference is provided, and then the estimated value is subtracted from the input space-time sample to eliminate the plane wave interference.

The second part is a beamformer, which is used to form a beam in the target direction.

The third part is time processing, including white filtering and matched filtering. Correlation interference fields are often encountered in the ocean, and better results can be obtained through optimization processing. The real-time optimization of passive sonar is realized.

In sonar, the actual optimal system must be able to constantly learn the surrounding environment, adjust the internal structural parameters at any time, and make the system performance as close to the best as possible according to some standards. Such a system is called an adaptive system. The best array processor realized by adaptive method is called adaptive beamformer. With the development of sonar signal processing technology, the output data rate of the receiver is getting higher and higher, and it is difficult for sonar personnel to identify the target and determine its parameters, which leads to the development of machine-aided detection and automatic detection technology. Although the theory of underwater acoustic signal processing is similar to that of radar, there are still many differences between them because of the complexity of underwater acoustic channel.