期刊名称：TELKOMNIKA (Telecommunication Computing Electronics and Control)
印刷版ISSN：2302-9293
出版年度：2017
卷号：15
期号：1
页码：314-321
DOI：10.12928/telkomnika.v15i1.4510
语种：English
出版社：Universitas Ahmad Dahlan
摘要：The biggest challenge in the underwater communication and target locating is to reduce the effect of underwater acoustic noise (UWAN). An experimental model is presented in this paper for the diurnal variability of UWAN of the acoustic underwater channel in tropical shallow water. Different segments of data are measured diurnally at various depths located in the Tanjung Balau, Johor, Malaysia. Most applications assume that the noise is white and Gaussian. However, the UWAN is not just thermal noise but a combination of turbulence, shipping and wind noises. Thus, it is appropriate to assume UWAN as colored rather than white noise. Site-specific noise, especially in shallow water often contains significant non-Gaussian components. The real-time noise segments are analyzed to determine the statistical properties such as power spectral density (PSD), autocorrelation function and probability density function (pdf). The results show the UWAN has a non-Gaussian pdf and is colored. Moreover, the difference in UWAN characteristics between day and night is studied and the noise power at night is found to be more than at the day time by around (3-8dB).
其他摘要：The biggest challenge in the underwater communication and target locating is to reduce the effect of underwater acoustic noise (UWAN). An experimental model is presented in this paper for the diurnal variability of UWAN of the acoustic underwater channel in tropical shallow water. Different segments of data are measured diurnally at various depths located in the Tanjung Balau, Johor, Malaysia. Most applications assume that the noise is white and Gaussian. However, the UWAN is not just thermal noise but a combination of turbulence, shipping and wind noises. Thus, it is appropriate to assume UWAN as colored rather than white noise. Site-specific noise, especially in shallow water often contains significant non-Gaussian components. The real-time noise segments are analyzed to determine the statistical properties such as power spectral density (PSD), autocorrelation function and probability density function (pdf). The results show the UWAN has a non-Gaussian pdf and is colored. Moreover, the difference in UWAN characteristics between day and night is studied and the noise power at night is found to be more than at the day time by around (3-8dB).
