Sequential filtering provides a suitable framework for estimating and updating the unknown parameters of system as data become available. The foundations sequential Bayesian with emphasis on practical issues are first reviewed covering both Kalman particle filter approaches. Filtering is demonstrated to be powerful estimation tool, employing prediction from previous estimates updates stemming physical statistical models that relate acoustic measurements parameters. Ocean applications then...

In this work, model-based methods are employed, along with machine learning techniques, to classify sediments in oceanic environments based on the geoacoustic properties of a two-layer seabed. Two different scenarios investigated. First, simple low-frequency case is set up, which acoustic field modeled normal modes. Four hypotheses made for seafloor sediment possibilities, and these explored using both various techniques matched-field approach. For most noise levels, latter has an inferior...

In the Hudson Canyon experiment, a sound source moved at constant depth in 73 m of water while transmitting four tonals. The signal was received on vertical array hydrophones that spanned column. data set from this experiment has become standard test case for studying tracking using matched field processing. As part process it important to first determine suitable environment model and demonstrate feasibility matched-field paper, we provide background original processing done accomplish...

For a sparsely observed acoustic field, Gaussian processes can predict densely sampled field on the array. The prediction quality depends choice of kernel and set hyperparameters. are applied to source localization in ocean combination with matched-field processing. Compared conventional processing, denser sampling predicted across array reduces ambiguity function sidelobes. As noise level increases, process-based processor has distinctly higher probability correct than due both denoising prediction.

Rapid voltage-controlled phase modulation of cw terahertz (THz) radiation is demonstrated. By transmitting an infrared beam through a lithium niobate modulator the THz radiation, which generated by photomixing two beams, can be directly modulated 2pi shift. The 100 kHz rate that demonstrated with this technique approximately 3 orders magnitude faster than what achieved mechanical scanning.

Gaussian processes (GPs) can capture correlation of the acoustic field at different depths in ocean. This feature is exploited this work for pre-processing data before these are employed source localization and environmental inversion using matched (MFI) an underwater waveguide. Via application GPs, denoised interpolated, generating densely populated fields virtual arrays, which then used as MFI. Replicas also computed receivers predictions made. The correlations among measurements distinct...

In this paper neural and statistical classifiers are applied to the problem of seafloor classification. The feature vectors used consist acoustic backscatter as a function angle incidence. Simulated is obtained by employing Helmholtz-Kirchhoff approximation properties bottom reverberation. These synthetic data initially train multilayer perceptrons then test them for their ability discriminate among signal returns produced seafloors with different roughness parameters. same also processed...

The use of machine learning (ML) in acoustics has received much attention the last decade. ML is unique that it can be applied to all areas acoustics. transformative potentials as extract statistically based new information about events observed acoustic data. Acoustic data provide scientific and engineering insight ranging from biology communications ocean Earth science. This special issue included 61 papers, illustrating very diverse applications

The stand-off imaging properties of a terahertz (THz) interferometric array are examined. For this application, the imaged object is in near-field region limit array. In region, spherical and circular architectures can compensate for distortions increase field view depth focus. Imaging THz point sources emphasized to demonstrate method compare theoretical predictions experimental performance.

Locating and tracking a source in an ocean environment estimating environmental parameters of sound propagation medium are critical tasks acoustics. Many approaches for both based on full field calculations which computationally intensive sensitive to assumptions the structure environment. Alternative methods that use only select features acoustic localization parameter estimation have been proposed. The focus this paper is development method extracts arrival times amplitudes distinct paths...

It has been previously shown using synthetic data that dispersion tracking with particle filtering can be used for sediment sound speed inversion. Here, is performed collected in the Gulf of Mexico and thickness water column depth estimation. In this experiment, propagates a long distance from source allows identification curves reflecting different group velocities modal frequencies within across modes. Although are noisy, tracked sequential Probability density functions three unknown...

Dispersion curves in ocean environments are accurately estimated from received signals through the extraction of instantaneous modal frequencies and corresponding arrival times for long-range propagation. The ultimate goal is to estimate sediment sound speed using extracted dispersion pattern. approach extends work previously conducted tracking with sequential filtering, improving on latter technique. state-space method that developed time-frequency information specific time instances relies...

Transmission of linearly frequency modulated pulses generates receptions at a vertical line array that can be cross correlated with the source signal to provide estimates oceanic waveguide impulse response. For short ranges, distinct path arrivals identified including direct, surface reflection, bottom and sediment reflection. Accurate estimation arrival times such paths is tightly related successful inversion for location water column depth sound speed and, subsequently, thickness. To...

Geoacoustic parameter estimation is presented as a non-linear regression problem where prediction performed using generalized additive models applied to features extracted from broadband acoustic time-series in machine learning framework. Qualitatively, it can be seen that signals have propagated different environments distinct structures: some cases, single mode identified, others, multiple modes seen; also distinguished by energy levels. Features are employed here comprise relative...

Multipath arrivals at a receiving sensor are frequently encountered in many signal-processing areas, including sonar, radar, and communication problems. In underwater acoustics, numerous approaches to source localization, geoacoustic inversion, tomography rely on accurate multipath arrival extraction. A novel method for estimation of time delays amplitudes with maximum posteriori (MAP) is presented here. MAP optimal if appropriate statistical models selected the data; implementation,...

A method is developed for the estimation of source location and sound speed in water column relying on linearization. The Jacobian matrix, necessary proposed linearization approach, includes derivatives with respect to empirical orthogonal function coefficients instead directly. First, inversion technique tested synthetic arrival times, using Gaussian distributions errors considered times. approach efficient, requiring a few iterations, produces accurate results. Probability densities...

In this paper, impulse response matching is proposed for source localization and environmental inversion. The ocean estimated using a cross-correlation procedure applied to data from the propagation of broadband pulse in shallow-water environment. Source geoacoustic parameter estimation are then performed through time-domain correlations between responses at spatially separated phones synthetic replica responses. method both temporally coherent. Parameter space search uses hierarchical...

Multiple source localization in underwater environments is approached within a matched-field processing framework. A maximum posteriori estimation method proposed that estimates location and spectral characteristics of multiple sources via Gibbs sampling. The facilitates weak which are typically masked by the presence strong interferers. performance evaluation study based on Monte Carlo simulations shows approach superior to simple coherent interference cancellation. also tested number...

Trace metals play important roles in regulating phytoplankton growth and could influence algal bloom development. Laboratory studies were conducted to evaluate the of environmentally relevant concentrations Cd, Cu, Ni, Zn on Aureococcus anophagefferens (brown tide) Results show that elevated Ni(2+) concentrations, e.g. those brown tide waters northeastern US, greatly stimulated A. (as compared control without Ni addition), yet, only low amounts dissolved sequestered, thus leaving excessive...

Matched-field processing is applied to source localization and detection of sound sources in the ocean. The spectrum included set unknown parameters estimated localization/detection process. Bayesian broadband (multi-tonal) incoherent coherent processors are developed, integrating estimation using a Gibbs sampler first evaluated via point estimates probability density functions obtained from synthetic signals. performance superior one both terms location spread. two also real data Hudson...

Matched-field processing is a method for inversion of the acoustic field utilizing its spatial coherence. In this work, matched-field processor introduced that incorporates coherence not only at single frequency but across frequencies as well. The new suitable multitonal sources and does require knowledge source spectrum which typically unavailable in passive estimation problems. A performance evaluation on localization under low signal-to-noise ratios shows significantly superior, certain...

Multilayer perceptrons trained with the backpropagation algorithm are tested in detection and classification tasks compared to optimal algorithms resulting from likelihood ratio tests. The focus is on problem of one M orthogonal signals a Gaussian noise environment, since both Bayesian detector classifier known for this can provide measure performance evaluation neural networks. Two basic situations considered: classification. For part, it was observed that signal-known-exactly case (M=1),...