Method of Spatially Correlated Wideband Ambient Noise Simulation for Underwater Acoustic Array

The spatial correlation characteristics of underwater ambient noise are critical factors for underwater acoustic array signal processing. A method for wideband underwater ambient noise simulation with specified spatial correlation characteristics is presented in this paper. Sub-band constant covariance matrices are first determined by frequency band decomposition with given correlation coefficient resolution base on the specified spatial correlation function. Cholesky decomposition of corresponding sub-band covariance matrix is then used to construct a linear transformation of independent band-limited white Gaussian random series in each frequency band to generate multiple correlative band-limited white Gaussian random time series, which are synthesized to form multichannel correlative wideband white Gaussian series. Finally, correlative wideband white Gaussian series of each channel are passed through an AR filter, which gives the output of multichannel colored ambient noise. Spatially correlated ambient noise of underwater acoustic arrays simulation with different frequency band and array shape are performed using the above method. The spectral characteristics of the simulation results and the correlation between the array elements match well with theoretical expectations.