Paper
24 May 2012 Empirical space-time statistical models for inhomogeneous acoustic propagation environments
Author Affiliations +
Abstract
When acoustic signals are subject to measurement over large distances or extended periods of time, the environmental conditions governing their propagation are unlikely to remain constant over the necessary spatial and temporal extents. Relative to a static environment, such inhomogeneities may result in severe signal distortion, such as non-linear warping, and can significantly degrade subsequent signal processing tasks such as classification and time-delay estimation. In this paper we 1) describe a set of experiments that were performed in order to collect space-time acoustic propagation data for empirical modeling, paying particular attention to important experimental design issues such as optimal sampling rates in the spatial domain, and 2) present a statistical two-dimensional model for inhomogeneous environments that describes the space-time distribution of acoustic propagation velocity governing low-frequency long-range propagation of aeroacoustic signals with long durations (several minutes). The model includes a deterministic component to model structured changes (e.g., increasing temperature during morning hours) and a stochastic component, specified by a two dimensional Gaussian random process, to capture correlated random deviations. Cram´er-Rao bounds are presented as a means of evaluating and optimizing sensor geometries for learning model parameters.
© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Joshua N. Ash "Empirical space-time statistical models for inhomogeneous acoustic propagation environments", Proc. SPIE 8388, Unattended Ground, Sea, and Air Sensor Technologies and Applications XIV, 83880G (24 May 2012); https://doi.org/10.1117/12.919762
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Data modeling

Sensors

Acoustics

Error analysis

Statistical analysis

Statistical modeling

Atmospheric modeling

Back to Top