Matched Field Processing Algorithm Research Articles

Extraction of surface-wave phase velocities from ambient noise in the presence of local noise sources based on matched-field processing

Passive surface wave methods have gained much attention from geophysics and civil engineering communities with the advantage of lower costs and deeper penetration depths compared to active seismic methods. The successful extraction of surface waves from ambient noise relies on the assumption of a homogeneous random distribution of noise sources. In urban environments, there are many strong noise sources located in localized areas (i.e., local sources) that violate the ideal noise distribution conditions, which causes errors in surface-wave dispersion extraction. To solve this problem, we propose a new passive surface wave analysis method in the presence of strong local noise sources (local-source passive surface wave analysis, LSPSW). We use the matched field processing (MFP) algorithm to capture noise source distributions and then extract reliable surface wave phase velocities. The MFP algorithm can accurately estimate the distribution of noise sources located either close to or far from the receiver array. Synthetic and field examples demonstrate the feasibility of using LSPSW to estimate phase velocities in complicated noise environments with a pseudo-linear array. Comparisons with other passive surface wave methods (seismic interferometry, the spatial autocorrelation method, pseudo-linear array analysis of passive surface waves based on beamforming) further illustrate the superiority of the MFP-based LSPSW method.

Underwater Target Localization Using Opportunistic Ship Noise Recorded on a Compact Hydrophone Array

In this research, a new application using broadband ship noise as a source-of-opportunity to estimate the scattering field from the underwater targets is reported. For this purpose, a field trial was conducted in collaboration with JASCO Applied Sciences at Duncan’s Cove, Canada in September 2020. A hydrophone array was deployed in the outbound shipping lane at a depth of approximately 71 m to collect broadband noise data from different ship types and effectively localize the underwater targets. In this experiment, a target was installed at a distance (93 m) from the hydrophone array at a depth of 25 m. In this study, a matched field processing (MFP) algorithm is utilized for localization. Different propagation models are presented using Green’s function to generate the replica signal; this includes normal modes in a shallow water waveguide, the Lloyd-mirror pattern for deep water, as well as the image model. We use the MFP algorithm with different types of underwater environment models and a proposed estimator to find the best match between the received signal and the replica signal. Finally, by applying the scatter function on the proposed multi-channel cross correlation coefficient time-frequency localization algorithm, the location of target is detected.

Induced seismicity during the 2012 Newberry EGS stimulation: Assessment of two advanced earthquake detection techniques at an EGS site

To improve the microearthquake detection capability during the 2012 Newberry enhanced geothermal system (EGS) demonstration, we applied two advanced earthquake detection algorithms to the continuous seismic data. These techniques have the power to detect microearthquakes within the seismic noise that would otherwise be missed using traditional techniques. We compare the efficacy of a matched field processing algorithm with a multi-channel waveform correlation algorithm. Both methods identified between 60%–80% more events, with each slightly outperforming the other in different seismic noise contexts. Incorporating either template-matching technique into a seismic processing routine could therefore effectively increase the sensitivity of the seismic network without changes to the existing hardware. Additionally, lowering the magnitude of completeness would significantly add to the number of events that could subsequently inform future decision making at the reservoir level.

Performance Evaluation of Acoustic Model-Based Blind Channel Estimation in Ocean Waveguides

Blind channel estimation (BCE) which does not require the knowledge of the source signals becomes increasingly important in various underwater acoustic applications. Unlike the conventional methods that utilize the statistical properties of the received signals, this paper proposes an acoustic model-based BCE method that exploits the properties of the acoustic propagation and the acoustic environmental information. This method is implemented by localizing the source with a matched-field processing algorithm first, then, uses the physical model of the ocean waveguides to determine the channel impulse response. Its performance is investigated by making comparisons with the mode-based artificial time reversal and the ray-based synthetic time reversal methods, using the array data measured from the SWellEx-96 experiment at four source-array ranges. Cross-correlation coefficient and normalized projection misalignment are adopted to evaluate the performance of the proposed method, and the simulation results for different multipath channels confirm the effectiveness of the proposed method.

Performance analysis of matched field processing localization with various line array configurations based on normal mode decomposition

Various line array configurations are evaluated for the source localization performance based on the analysis of mode decomposition matrix in this paper. The guideline of array shape design focuses on improving the localization performance of matched filed processing, meanwhile reducing the difficulty of deploying equipment in practical experiments. In the shallow water environment, when the environment is well known, the source localization result can be obtained by matched field processing algorithms effectively, but the source localization performance is affected by the array parameters, such as array length, the number of sensors, and the configurations of various horizontal and vertical line arrays. The modal decomposition method provides a useful insight into the questions of how many modes are needed and how to design the array to resolve the modes. Therefore, the method of utilizing a normal mode acoustic propagation model to decompose mode is proposed by vertical line array, horizontal line array and combined array respectively. Then we can evaluate the source localization performance of various line array configurations by studying the characteristic of normal mode decomposition matrix, thus establishing a qualitative or even quantitative relationship between each other. The more the normal mode decomposition matrix tends to be diagonalized, the better performance of line array localization will be obtained. Simulation results show that the localization performance of matched field processing with the combined arrays will be severely degraded when the mode amplitudes cannot be accurately deduced by one of the sub-arrays. Considering the requirements for the practical experiments and various environments, the source localization performance of short vertical line array and combined array are mainly discussed in this paper. The combined array can increase the azimuth and depth information of the source and realize three-dimensional target detection while the vertical array provides range-depth information and the horizontal array provides bearing information. Simulation result indicates that the design guidelines based on the normal mode decomposition are appropriate for arrays employed for matched filed processing. Meanwhile, the combined arrays perform better than the short vertical array, which is benefited by the horizontal array's suppressing the side lobes, which leads the ratio of peak to sidelobe to increase, and thus improving the location accuracy. The values of localization accuracy of combined arrays are all above 90% according to the simulation experiment. Take the practical application into account, the combined array is undoubtedly a compromise choice for the localization performance and the test complexity.

Motion planning based on matched field processing with Bayesian filtering

Matched Field Processing (MFP) has often been used with stationary arrays as a method for developing source location information, but potentially with large amounts of uncertainty. This has been overcome in some cases by using Bayesian Filtering to improve estimates over time of the source location by treating this quantity as a random variable. Here we show how these methods can be applied to an autonomous vehicle for motion planning. The vehicle moving through an acoustic field can provide spatially separated data that can be used in an MFP algorithm to generate an estimated source location, which can be integrated utilizing Bayesian Filtering. Motion planning methods are shown which aim to move the vehicle to the maxima of this probability distribution. This method is demonstrated using ocean acoustic simulations. [Work supported by Raytheon Advanced Studies Fellowship.]

Adaptive Matched Field Processing for Source Localization Using Improved Diagonal Loading Algorithm

Matched field processing (MFP) has been a method widely applied for shallow underwater target localization, which is a critical issue in underwater acoustic. To enhance the efficiency of conventional MFP methods, different adaptive MFP algorithms have been developed; the white noise constraints (WNC) MFP or diagonal loading (DL) algorithm is such a typical one. The WNC or DL one has been considered to be the most desirable method because it is more robust to environment mismatch in practical in comparison with the minimum-variance distortionless response MFP algorithm, a popular high-resolution method. Although having exceptional ability to localize underwater sources in mismatch scenarios, the DL method has still been not reach high resolution in certain cases. In the paper, we proposed an adaptive method known as improved diagonal loading algorithm to make an increase in the resolution and the peak background rate in the ambiguity surface of source localization results in comparison with DL one. The proposed algorithm works by adding one more parameter that is adjusted in the steering vector of the DL algorithm. The simulation results show that the new algorithm attains better beamforming performance in terms of high resolution than the existing adaptive MFP algorithms in the case of environmental mismatch caused by noise effects and the limitation of the snapshots.

Use of polynomial chaos expansions in the design of uncertainty-robust matched field processing algorithms

The polynomial chaos (PC) expansion method has been recently applied to estimating the statistical properties of underwater acoustic propagation in the presence of environmental uncertainty. Here we use PC estimates of the field covariance structure to design uncertainty robust match field processing (MFP) weights using Krolik’s minimum variance beamformer with sound-speed perturbation constraints (MV-SPC) [J. Acoust. Soc. Am. 92 (3), 1408–1419 (1992)]. The idea behind the MV-SPC beamformer is to realize much of the high sidelobe rejection of MV processors in environments with environmental variability by opening up the signal model to include uncertainty effects. Here we compare the performance of MV-SPC designed with an adiabatic signal uncertainty model to the same beamformer designed with the PC signal model for realizations of the signal vector obtained with fully coupled propagation models, with results showing the superiority of the PC approach. [Work supported by ONR.]

FAULT-TOLERANT MATCHED-FIELD PROCESSING IN THE PRESENCE OF ELEMENT FAILURES

In the highly cluttered undersea environment, sonar array systems require enhanced acoustic signal processing algorithms and sophisticated architectures in order to meet dependability and real-time mission requirements. The probability of hydrophone and processing element failures is very high in such severe operating environments. Adaptive matched-field processing (MFP) algorithms localize sources accurately with moderate levels of signal-to-noise ratio (SNR) and precise knowledge about environments by employing full-wave acoustic propagation models. However, they highly distort output beam patterns with significant increase of sidelobes in the presence of environmental mismatches and element failures. These problems make the development of advanced fault-tolerant signal processing algorithms imperative to tolerate the element failures in cases where replacement of defective elements is impossible or impractical. In this paper, three fault-tolerant MFP algorithms are presented to compensate for the performance degradation generated by the inherent failure characteristics of vertical line arrays. The beamforming performance and computational complexities for these fault-tolerant algorithms are analyzed in terms of the number of faulty elements, their positions in the array, and SNRs. The simulation results demonstrate that these fault-tolerant techniques provide a feasible solution for real-time and highly reliable beamforming implementation on sonar array systems.

FAULT-TOLERANT PARALLEL ALGORITHMS FOR ADAPTIVE MATCHED-FIELD PROCESSING ON DISTRIBUTED ARRAY SYSTEMS

Continuous innovations in adaptive matched-field processing (MFP) algorithms have presented significant increases in computational complexity and resource requirements that make development and use of advanced parallel processing techniques imperative. In real-time sonar systems operating in severe underwater environments, there is a high likelihood of some part of systems exhibiting defective behavior, resulting in loss of critical network, processor, and sensor elements, and degradation in beam power pattern. Such real-time sonar systems require high reliability to overcome these challenging problems. In this paper, efficient fault-tolerant parallel algorithms based on coarse-grained domain decomposition methods are developed in order to meet real-time and reliability requirements on distributed array systems in the presence of processor and sensor element failures. The performance of the fault-tolerant parallel algorithms is experimentally analyzed in terms of beamforming performance, computation time, speedup, and parallel efficiency on a distributed testbed. The performance results demonstrate that these fault-tolerant parallel algorithms can provide real-time, scalable, lightweight, and fault-tolerant implementations for adaptive MFP algorithms on distributed array systems.

PARALLEL ALGORITHMS FOR ADAPTIVE MATCHED-FIELD PROCESSING ON DISTRIBUTED ARRAY SYSTEMS

Matched-field processing (MFP) localizes sources more accurately than plane-wave beamforming by employing full-wave acoustic propagation models for the cluttered ocean environment. The minimum variance distortionless response MFP (MVDR–MFP) algorithm incorporates the MVDR technique into the MFP algorithm to enhance beamforming performance. Such an adaptive MFP algorithm involves intensive computational and memory requirements due to its complex acoustic model and environmental adaptation. The real-time implementation of adaptive MFP algorithms for large surveillance areas presents a serious computational challenge where high-performance embedded computing and parallel processing may be required to meet real-time constraints. In this paper, three parallel algorithms based on domain decomposition techniques are presented for the MVDR–MFP algorithm on distributed array systems. The parallel performance factors in terms of execution times, communication times, parallel efficiencies, and memory capacities are examined on three potential distributed systems including two types of digital signal processor arrays and a cluster of personal computers. The performance results demonstrate that these parallel algorithms provide a feasible solution for real-time, scalable, and cost-effective adaptive beamforming on embedded, distributed array systems.

Reduced beamset adaptive matched field processing

Matched field processing (MFP) offers the possibility of improved towed array performance at endfire through range/depth discrimination of contacts. One challenge is that arrays with limited vertical aperture can often resolve only a small number of multipath arrivals. This paper explores ways to capture the array resolution by re-parametrizing the set of MFP replicas. A reduced beamset can be created by performing a singular value decomposition on the MFP replica set. Alternatively, clustering techniques can be used to generate MFP cell families, or regions of similar response. These parametrizations are applied to adaptive MFP algorithms to show speed and performance gains. The use of cell families/regions instead of individual MFP cells also provides a framework for increasing the robustness of MFP by defocusing the MFP beamforming operation. The techniques are demonstrated for shallow-water towed array scenarios. [Work sponsored by DARPA-ATO under Air Force Contract No. F19628-00-C-0002. Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the Department of Defense. Approved for Public Release, Distribution Unlimited.]

Information theory for matched field processing: preliminary results

Matched field processing (MFP) has traditionally been considered as an estimation problem in which the goal is to use the observed pressure field and an acoustic propagation model to estimate an unknown source location with the smallest possible variance. An alternative perspective is to divide the search region into cells whose size is dictated by logistical constraints, and attempt to assign the source to one of these cells. In this perspective, the goal of MFP is to minimize the probability of error in assigning the source to a grid cell, rather than minimizing the variance of the source location estimate. This perspective leads one to consider MFP as a communication problem in which the source transmits its location to the receiver array. Information theory provides techniques which can bound the amount of information available to the receiver in this problem. The amount of information observable has implications for the performance limits achievable by any MFP algorithm. Additionally, this perspective leads to methods of designing arrays for MFP which maximize the average information observed over the search space. [Work supported by ONR Young Investigator Program.]

Comparison of wide-band matched field processing algorithms

There are several ways to realize a wide-band MFP process. The data may be processed in the frequency domain, one frequency at a time, and then incoherently or coherently integrated over the band. Alternatively, the data may be processed in the time domain, by matching modeled correlations (e.g., calculated using a ray trace model) with measured correlations. Auto correlations at a single sensor or interelement cross correlations can be used. We compare the frequency-domain and time-domain approaches with respect to array configuration requirements, sensitivity to array element location errors, and sensitivity to environmental mismatch. We demonstrate our findings using simulated and experiment data.

Source localization on a ‘‘fuzzy structure’’ beam using matched-field processing

Matched-field processing (MFP), a technique originally developed by the underwater acoustic community [Baggeroer, Kuperman, and Mikhalevsky, IEEE J. Ocean Eng. 18, 401–424 (1993)] for source localization, is used to localize the source of excitation on a flexurally vibrating beam to which is attached a ‘‘fuzzy structure (FS).’’ The FS, for this study, consists of a set of flexurally vibrating strips, of randomly varying lengths, attached at uniformly spaced points along the length of the beam. MFP differs from classical inverse methods in that it does not require an inversion of the system model. Instead it correlates or ‘‘matches’’ observed signals with solutions of the equations of motion for a range of excitation points using a variety of linear and nonlinear methods. In this paper several different MFP algorithms are used and compared where the observed or simulated data is measured both globally, along the entire length of the beam, or locally over a limited region of the structure. Successful localization is found to be dependent on frequency. For lightly damped structures, the MFP localization results become ambiguous at the resonance frequencies of the structure. [Work supported by ILIR Program CDNSWC.]

Computationally efficient, robust algorithm for matched field processing

The matched field processing (MFP) localization performance of very low frequency (VLF) arrays operated in the deep ocean basins appears to be limited more by uncertainty in the sound velocity profile (SVP) than by low signal-to-noise ratio (SNR). We present a new robust variation of MFP designed to be less sensitive to velocity error in weakly inhomogeneous environments. We analyze the computational requirements of this and other MFP algorithms. When either the search volume is large of the acoustic array is large, computational efficiency is an issue. We present an efficient MFP implementation for the conventional MFP algorithm and our robust algorithm. We show that parallel implementation of these algorithms may allow real-time performance.

Matched-field source localization in a random ocean

This paper is devoted to the statistical approach to matched-field processing (MFP). In Sec. I, a statistical model of the Green’s function of a random oceanic waveguide is described using the mean field and the second moment. In Sec. 2, using Bayesian approach, optimum target detection and localization algorithms matched to a deterministic and a stochastic ocean are derived. First, the optimum decisive statistic (log-likelihood ratio) is derived for a deterministic ocean. In the general case, ocean parameter fluctuations make the Green’s function random and so we need to average the log-likelihood ratio. The MFP algorithms for a random ocean waveguide developed on the basis of the Bayesian approach are less precise than the algorithms for a deterministic oceanic waveguide model, but they are more robust. In Sec. 3 a comparative analysis of the proposed MFP methods and the known suboptimum MFP techniques is carried out. The ambiguity surfaces for the proposed algorithms, Bartlett, Minimum Variance (MV), MV with neighboring location constraints, and MV with sound-speed perturbation constraints methods are presented. The proposed MFP method outperformed other algorithms under random ocean conditions, giving the least bias error and the lowest sidelobes.

Shallow water beamforming with small aperture, horizontal, towed arrays

This paper shows that the performance of plane-wave beamformers is significantly degraded in shallow water for horizontal arrays, even for short aperture arrays. Horizontal towed arrays must have short apertures for practical reasons, and the plane-wave beamformer needs to be augmented to optimize performance in shallow water. Matched-field processing (MFP) is a beamforming technique which allows for the detection and localization of an acoustic source in range and depth using passive sonar. The performance of MFP to augment plane-wave beamforming in shallow water environments using a short aperture horizontal line array is investigated. It is realized that MFP performance is enhanced when vertical arrays are used in deep water, or large aperture horizontal arrays are available in shallow water. However, this paper addresses the realistic case of operational, tactical horizontal arrays which are practical to deploy in shallow water. This paper attempts to integrate inverse beamforming (IBF) [A. H. Nuttall and J. H. Wilson, J. Acoust. Soc. Am. 90, 2004–2019 (1991); J. H. Wilson, J. Acoust. Soc. Am. 98, 3250–3261 (1995)], a plane-wave beamformer, with MFP so that the strengths of each are utilized. Results of test cases using synthetic data are presented to evaluate the effects of frequency, bottom type, relative target bearing, and sound-speed profile on MFP and IBF performance. IBF consists of three algorithms: the Fourier integral method (FIM) beamformer; the eight nearest neighbor peak picker (ENNPP); and a sophisticated M of N tracker. It is shown that MFP performance is significantly better than plane-wave beamforming for higher frequencies, for more reflective bottom types in shallow water, and for relative target bearings away from the broadside beams. IBF, or plane-wave beamforming, performs well at very low frequencies (VLF), in mud/silt-clay bottoms, and at beams near broadside. It is also shown that the performance of any plane-wave beamformer, including IBF, is severely degraded in shallow water for relative bearings away from the broadside beams, particularly at higher frequencies. Finally, the MFP algorithm in conjunction with the ENNPP and M of N tracker are shown to perform well in the real, deep water ocean environment. No appropriate shallow water measured data is available to evaluate the integrated IBF/MFP algorithm in shallow water.

Estimation and detection issues in matched-field processing

A conceptual framework in which the model-based, space-time acoustic signal processing procedure known as matched field processing (MFP) can be handled in a consistent manner is established. A framework for strong-signal MFP based on standard statistical estimation theory, in which MFP is regarded as essentially an estimation problem in the strong-signal regime, is developed. In the weak-signal case, the necessary requirement of detection dictates that MFP then be considered a joint detection-estimation task. It is demonstrated that, generally, MFP is essentially a space-time processing problem rather than simply an array processing (spatial processing only) procedure. An overview of the processing schemes used to date in MFP is given, showing how these methods relate to the optimal space-time structure. Weak-signal detection and estimation scenarios relevant to MFP are then noted. Present methods for dealing with the inherent instability of MFP algorithms (mismatch) are discussed. The current status of MFP is summarized, and recommendations for future research are made.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Performance stability of high-resolution matched-field processors to sound-speed mismatch in a shallow-water environment

The effects of variations in water sound-speed parameters on the performance of four matched-field processing algorithms (Bartlett, maximum-likelihood, sector-focusing, and multiple constraint) have been investigated. The SNAP propagation model was used to generate the replica acoustic pressure field for a shallow water channel with a depth variable sound-speed profile typical of a midlatitude summer environment. It was also used to simulate a ‘‘detected’’ field due to an acoustic source. These were then correlated using the four algorithms for selected degrees of mismatch of the water sound-speed profile. The maximum likelihood and multiple constraint estimators achieved good peak resolution and high accuracy for small degrees of mismatch. The maximum likelihood estimator deteriorated quickly as the mismatch increased. The multiple constraint estimator performed significantly better, but both failed to give correct location estimates at the higher mismatch values. In contrast, the sector focusing estimator continued to give accurate location estimates over the whole range of sound-speed mismatch, with slightly less peak resolution at small mismatch values. It also performed significantly better overall than the Bartlett estimator, which was found to give accurate location estimates over the whole range of mismatch values used, but with poor peak resolution against a background containing many high sidelobes.