**6. Characterization and denoising of ambient noise**

Characterization of ambient noise had become the epicenter during World War II. The need to develop acoustic sensor and systems had become of prime importance for processing different sound signatures of submarines [1]. This emphasized the need of noise characterization, and with this interest, many engineers and researchers started working with sensors and sonar for signal reception and processing.

After characterization of ambient noise, data analysis comes, which is essential for research and pragmatic approach. It serves two objectives: determining parameters that are needed for constructing necessary model and thereafter authenticating the constructed model.

Irrespective of physical measurements or quantitative modeling, data has one or more issues to deal with: (1) shorter total data span, (2) nonstationary data, and (3) data representing nonlinear processes. Here the first two issues are interrelated, for section of data shorter than the longest timescale of a stationary process may appear to be nonstationary. There are limited options available to analyze this kind of data (http://rcada.ncu.edu.tw/).

The recovery of the signal buried in ambient noise is of importance for the target's signal detection, recognition, and classification at low signal-to-noise ratio (Tieshuang et al. [14]).

## **7. Conclusion**

In this chapter we have done a detailed study about underwater ambient noise, various sources that produce the noise, and shallow water ambient noise. In science and engineering, noise is defined as an unwanted energy. Noise is relentless that blots out or reduces the intensity of a signal. Noise can be a disturbance that may be internally generated because of the process itself or due to some irresolvable issues or may be due to intermittent and local instability. This can be produced by various recording systems and different types of sensors. It is important to analyze the data differences that would exist between actual data and the noise so that attempts can be done to remove this noise.

**159**

**Author details**

Vijaya Baskar Veeriayan1

and drvrajen@gmail.com

provided the original work is properly cited.

\* and Rajendran V.<sup>2</sup>

\*Address all correspondence to: vijayabaskar.ece@sathyabama.ac.in

2 Department of ECE, VELS University, Chennai, India

\*

1 School of EEE, Sathyabama Institute of Science and Technology, Chennai, India

© 2020 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/3.0), which permits unrestricted use, distribution, and reproduction in any medium,

*Underwater Ambient Noise*

*DOI: http://dx.doi.org/10.5772/intechopen.93057*

*Underwater Ambient Noise DOI: http://dx.doi.org/10.5772/intechopen.93057*

*Noise and Environment*

processing.

ing the constructed model.

(Tieshuang et al. [14]).

be done to remove this noise.

**7. Conclusion**

of data (http://rcada.ncu.edu.tw/).

Ambient noise is more in shallow water as the noise is pinned between the seafloor and the surface of the ocean. In shallow water (depth of 5–200 m), acoustic systems like sonar, echo sounder, and sub-bottom profiler suffer a huge loss due to ambient noise [13]. Wind speed in shallow water may vary from 0.5 m/s to 300 m/s;

Characterization of ambient noise had become the epicenter during World War II. The need to develop acoustic sensor and systems had become of prime importance for processing different sound signatures of submarines [1]. This emphasized the need of noise characterization, and with this interest, many engineers and researchers started working with sensors and sonar for signal reception and

After characterization of ambient noise, data analysis comes, which is essential for research and pragmatic approach. It serves two objectives: determining parameters that are needed for constructing necessary model and thereafter authenticat-

Irrespective of physical measurements or quantitative modeling, data has one or more issues to deal with: (1) shorter total data span, (2) nonstationary data, and (3) data representing nonlinear processes. Here the first two issues are interrelated, for section of data shorter than the longest timescale of a stationary process may appear to be nonstationary. There are limited options available to analyze this kind

The recovery of the signal buried in ambient noise is of importance for the target's signal detection, recognition, and classification at low signal-to-noise ratio

In this chapter we have done a detailed study about underwater ambient noise, various sources that produce the noise, and shallow water ambient noise. In science and engineering, noise is defined as an unwanted energy. Noise is relentless that blots out or reduces the intensity of a signal. Noise can be a disturbance that may be internally generated because of the process itself or due to some irresolvable issues or may be due to intermittent and local instability. This can be produced by various recording systems and different types of sensors. It is important to analyze the data differences that would exist between actual data and the noise so that attempts can

hence it causes disturbance to surface water, and these disturbances thereby propagate towards the seabed. Due to these factors the SNR value reduces. This poses a greater challenge to developing a method or algorithm to improve the SNR value. Hence, to improve the effectiveness of underwater acoustic instruments, the

measurement and characterization of ambient noise is more significant.

**6. Characterization and denoising of ambient noise**

**158**
