Physics-based Signal Processing Research Articles

Reflections on Lisa Zurk as an Educator

Lisa Zurk was my advisor and mentor through both my masters and doctoral degree programs. I had the pleasure of working with her beginning in 2008; first as her graduate student, then in my subsequent professional career while she was a program manager in the Strategic Technology Office at DARPA. I consider myself extremely fortunate to have been part of the Northwest Electromagnetics and Acoustics Research Laboratory (NEAR-Lab), which she founded at Portland State University in 2005. Under her guidance and leadership, the group became a diverse community of tight-knit students, faculty, and other researchers; united in our passion for research in terahertz imaging and spectroscopy, underwater acoustics, and physics-based signal processing. My experience there, and that of my fellow students, paved the way for our future careers in science and engineering in industry, academia, and everywhere in between. Today, NEAR-Lab alumni can be found at top tech companies such as Apple, Boeing, Intel, Tektronix, and Metron; nationally-renowned research organizations like The Johns Hopkins University Applied Physics Laboratory, MIT Lincoln Laboratory, and the Pacific Northwest National Laboratory; and even internationally at the University of Victoria, British Columbia, and JASCO Applied Sciences. The success of her students speaks volumes about her impact on the generation of innovators that followed her. This talk will reflect on Dr. Zurk as an educator and will present some of the work she coauthored and published with her students.

A survey of Lisa Zurk's contributions to physics-based signal processing in underwater acoustics

The incorporation and exploitation of the physics of underwater acoustic propagation (including reflection and scattering) in signal and information processing can improve the performance of detection, classification, localization and tracking (DCLT) in sonar systems. This presentation will survey several areas of Prof. Lisa Zurk's research from this perspective of physics-based signal processing. Contributions are found in the areas of (i) assessing the impact of propagation through a non-stationary underwater environment on the performance of signal processing algorithms, (ii) adapting signal processing algorithms to be more robust to these environmental effects, and (iii) the exploitation of specific characteristics of the propagation channel in novel approaches to solving the above inferential objectives. Of these areas, identifying consistently observable propagation phenomena that can be exploited to improve DCLT performance is the most essential and yet the most difficult because it requires balancing just enough accuracy in the physical model with a simple enough representation to develop robust signal and information processing algorithms. Impressively, Prof. Zurk was quite successful in this crucial area of physics-based signal processing.

Integration of a Novel Knowledge-Guided Loss Function With an Architecturally Explainable Network for Machine Degradation Modeling

The suitability of a health index (HI) is significant to improve machine diagnostic and prognostic performance. A life cycle HI can be applied to the detection of incipient faults and determination of the first prediction time, and then its monotonic curve after FPT can be used as an immediate variable for prognostic analysis. Although existing data-driven methodologies for HI construction have considerable nonlinear mapping, they lack transparency and interpretability for subsequent studies and analysis. Moreover, these HIs have some unknown burrs and fluctuations. To overcome these limitations, an architecturally explainable network for machine degradation modeling is proposed in this paper. The first three hidden layers are reformulated from three advanced signal processing technologies including Hilbert transform, squared envelope, and Fourier transform to map temporal signals in the time domain into demodulated signals in the squared envelope spectrum domain so as to directly link the proposed network with fault frequencies and their harmonics in the frequency domain. The fourth hidden layer consists of a full connection layer with one hidden node and a novel knowledge-guided loss function specifically designed for exploring and extracting informative degradation features. Case studies are implemented to verify the proposed methodology and show that the proposed HIs have more stable and robust performance compared to state of the art methods. Another superiority of the proposed network lies in its full transparency and interpretability by effectively integrating physics-based signal processing technologies with an interpretable artificial network structure for machine degradation modeling.

A Physics-Based Signal Processing Approach for Noninvasive Ultrasonic Characterization of Multiphase Oil-Water-Gas Flows in a Pipe.

A signal processing technique is presented for determining the composition of multiphase oil-water-gas flow in a pipe using noninvasive ultrasonic speed of sound measurements from a transmitter-receiver pair bonded to diametrically opposite sides of a pipe. A linear chirp excitation is used to send broadband ultrasonic energy that propagates in two paths from transmitter to receiver such as: 1) a wave through the pipe wall and then the multiphase mixture and 2) ultrasonic guided waves along the pipe wall in the circumferential direction. As the ultrasonic attenuation of the multiphase mixture increases, the amplitude of the signal through the fluid mixture decreases relative to that of circumferential guided waves, making it difficult to determine the time of arrival of the fluid-path signal and, hence, the speed of sound in the mixture. The proposed signal processing technique overcomes this challenge by using: 1) a guided wave subtraction approach to suppress the strength of guided wave signals relative to the fluid-path signal and 2) a Gaussian reconstruction approach for synthetic enhancement of the fluid-path signal by output signal reconstruction at frequencies corresponding to peak transmission of ultrasonic energy. The efficacy of the technique is demonstrated using experiments carried out in a field-scale flow loop with varying compositions of oil-water-gas mixtures. It is shown that the proposed approach can enhance the signal detectability by approximately 20 dB in comparison with the traditional approach that does not utilize guided wave subtraction and also improves the gas tolerance of composition measurements up to 20% in gas volume fraction.

Nonlinear time-warping made simple: A step-by-step tutorial on underwater acoustic modal separation with a single hydrophone.

Classical ocean acoustic experiments involve the use of synchronized arrays of sensors. However, the need to cover large areas and/or the use of small robotic platforms has evoked interest in single-hydrophone processing methods for localizing a source or characterizing the propagation environment. One such processing method is "warping," a non-linear, physics-based signal processing tool dedicated to decomposing multipath features of low-frequency transient signals (frequency f < 500 Hz), after their propagation through shallow water (depth D < 200 m) and their reception on a distant single hydrophone (range r > 1 km). Since its introduction to the underwater acoustics community in 2010, warping has been adopted in the ocean acoustics literature, mostly as a pre-processing method for single receiver geoacoustic inversion. Warping also has potential applications in other specialties, including bioacoustics; however, the technique can be daunting to many potential users unfamiliar with its intricacies. Consequently, this tutorial article covers basic warping theory, presents simulation examples, and provides practical experimental strategies. Accompanying supplementary material provides matlab code and simulated and experimental datasets for easy implementation of warping on both impulsive and frequency-modulated signals from both biotic and man-made sources. This combined material should provide interested readers with user-friendly resources for implementing warping methods into their own research.

Striation Processing of Data From the 2013 Target and Reverberation Experiment (TREX13)

In active sonar processing, the discrimination of the target versus clutter can be a significant challenge. Whereas conventional processing often uses target kinematics to limit the possible target tracks, it has recently been shown that frequency domain information (based on the waveguide invariant principle) can also be incorporated to further limit the possible target tracks in an environmentally robust fashion. This paper presents physics-based signal processing methods to extract information about the target track from striations in a target spectrogram formed from the echo spectra at each active sonar pulse repetition interval. The target tracking information is formulated as a post-track likeliness statistic that is extracted from the sonar data with image processing techniques. Results are demonstrated with shallow water sonar data collected during the 2013 Target and Reverberation Experiment (TREX13). It is expected that the physics-based signal processing algorithms discussed here will provide enhanced clutter rejection and improve tracking performance.

Striation processing of continuously active sonar data

Previous work by the authors has shown that striations associated with the waveguide invariant can provide a new physics-based tracking constraint that improves tracking performance in pulsed active sonar data. A necessary pre-processing step is to isolate the target return from the direct arrival and other target echoes. Striation-based beamforming can be used as an optional pre-processing step to preserve the multipath structure in the beamformer output of a horizontal line array. Recent interest in continuously active sonar (CAS) processing has motivated research to determine how striations might be similarly exploited for tracking when the sonar is operated with a high duty cycle. In CAS signals, the direct arrival is almost always overlapping in time with the target echo return and the standard time-gating technique cannot isolate the echo return. In this paper, a time-dependent filter is introduced to isolate and extract the CAS echo spectrum. The conditions in which striation-based beamforming should be applied to preserve the striation structure in the beam output are also explored. The physics-based signal processing algorithms discussed here are expected to provide valuable information about the target’s multipath structure, which can be exploited in an extended Kalman filter for enhanced tracker performance.

Statistical and Adaptive Signal Processing For UXO Discrimination for Next-Generation Sensor Data

Until recently, detection algorithms could not reliably distinguish between buried UXO and clutter, leading to many false alarms. Over the last several years modern geophysical techniques have been developed that merge more sophisticated sensors, underlying physical models, statistical signal processing algorithms, and adaptive training techniques. These new approaches have dramatically reduced false alarm rates, although for the most part they have been applied to data collected at sites with relatively benign topology and anomaly densities. On more challenging sites, performance of even these more modern discrimination approaches is still quite poor. As a result, efforts are underway to develop a new generation of UXO sensors that will produce data streams of multi-axis vector or gradiometric measurements, for which optimal processing has not yet been carefully considered or developed. We describe a research program to address this processing gap, employing a synergistic use of advanced phenomenologicalmodeling and signal-processing algorithms. The key foci of the program are (1) development of new physics-based signal processing approaches applicable to the problem in which vector data is available from such sensors; and (2) development of the theory of optimal experiments to guide the optimal design and deployment of the new sensor modalities. Here, we present initial results using simulated data obtained with our phenomenological models that indicate that optimal processing of features extracted from multi-axis EMI data can provide substantial improvements in discrimination performance over processing of features extracted from single-axis data.