Impact of uncertain environmental knowledge on the shallow-water transfer function

Abstract

It is assumed that the acoustic channel can be modeled as a linear time-invariant space-variant filter. In this case, from linear systems theory, it is known that the filter output, that is the predicted source signal replica at the sensor location, is formed by the convolution of the source signal with the filter impulse response function. The acoustic channel impulse response function is a function of the source location, the sensor location and the environmental

parameters. In this analysis we use the channel transfer function, the fourier transform of the channel impulse response function, as a function of frequency to evaluate the effect of environmental parameter uncertainty. In the course

of this work we establish a measure, referred to as the 'transfer function error measure' which provides an estimate of the average error for the channel transfer function due to environmental parameter uncertainty as a function

of receiver depth. This error measure, which is also a function of the source location, is used to characterize the sensitivity of propagation model-based array processors to uncertainty in environmental knowledge such as water sound speed, bottom sound speed, bottom attenuation and channel depth. For simple canonical shallow-water channels, winter and summer profiles, results are presented which illustrate the effect of uncertain environmental knowledge on the transfer function error measure.