Hot topics of signal processing in acoustics.

Abstract

Signal processing is used to some extent in all areas of acoustics, such as extracting relevant information from acoustic measurements made either in the laboratory or in the field, processing signals and/or synthesizing data to cope with demanding tasks raised in acoustics. Techniques range from simple classical approaches based on Fourier transforms and Gaussian noise, to sophisticated model-based techniques that incorporate physical/parametric models of the acoustical system. In this paper we highlight new approaches to signal processing that could be applied to a broad variety of acoustical problems. These include coded signals for architectural-acoustics, acoustical communications, and medium characterization, Bayesian methods for room acoustics, physical acoustics, and underwater acoustics including highly nonlinear problems with non-Gaussian noise, and extensions to the familiar Kalman filtering to nonlinear models. Examples of each approach will be shown that illustrate the advantages and disadvantages of each technique. Additional topics may be discussed as time allows.