Model-based Signal Processing Research Articles

A Novel Method for Automated Estimation of Effective Parameters of Complex Auditory Brainstem Response: Adaptive Processing Based on the Correntropy Concept

Objectives: Automated Auditory Brainstem Responses (ABR) peak detection is a novel technique to facilitate the measurement of neural synchrony along the auditory pathway through the brainstem. Analyzing the location of the peaks in these signals and the time interval between them may be utilized either for analyzing the hearing process or detecting peripheral and central lesions in the human hearing system. Methods: In this paper, model-based signal processing is proposed to estimate the effective parameters of ABR signals. In this process, the biological parameters of the signal are assessed by utilizing a Finite Impulse Response (FIR) adaptive filter in which its adaptation procedure is performed based on the correntropy concept. The proposed method is applied on a set of ABR signals recorded in response to three stimuli of /da/, /ba/, and /ga/, and then its performances are compared with an existing state-of-the-art technique. Results: The results show that the proposed method can significantly increase the accuracy of estimating the parameters in stable stimulations (/da/, /ba/) for major positive and negative peaks. This improvement is more significant (up to 2-3 times) for /ba/ stimulus and especially in major positive peaks. However, in other peaks, the improvements also occurred in smaller amounts. However, for unstable stimuli (/ga/), no significant improvement was achieved. Discussion: Increasing the accuracy performance of the proposed method for detecting the stable stimuli (while its performance remains unchanged) for detecting unstable stimuli indicates its effectiveness in automated clinical analysis of ABR signals.

Model-Based Prediction and Optimal Control of Pandemics by Non-Pharmaceutical Interventions

A model-based signal processing framework is proposed for pandemic trend forecasting and control, by using non-pharmaceutical interventions (NPI) at regional and country levels worldwide. The control objective is to prescribe quantifiable NPI strategies at different levels of stringency, which balance between human factors (such as new cases and death rates) and cost of intervention per region/country. Due to infrastructural disparities and differences in priorities of regions and countries, strategists are given the flexibility to weight between different NPIs and to select the desired balance between the human factor and overall NPI cost. The proposed framework is based on a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">finite-horizon optimal control</i> (FHOC) formulation of the bi-objective problem and the FHOC is numerically solved by using an ad hoc <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">extended Kalman filtering/smoothing</i> framework for optimal NPI estimation and pandemic trend forecasting. The algorithm enables strategists to select the desired balance between the human factor and NPI cost with a set of weights and parameters. The parameters of the model are partially selected by epidemiological facts from COVID-19 studies, and partially trained by using machine learning techniques. The developed algorithm is applied on ground truth data from the Oxford COVID-19 Government Response Tracker project, which has categorized and quantified the regional responses to the pandemic for more than 300 countries and regions worldwide, since January 2020. The dataset was used for NPI-based prediction and prescription during the XPRIZE Pandemic Response Challenge.

A Model-Based Analysis of Capacitive Flow Metering for Pneumatic Conveying Systems: A Comparison between Calibration-Based and Tomographic Approaches.

Pneumatic conveying is a standard transportation technique for bulk materials in various industrial fields. Flow metering is crucial for the efficient and reliable operation of such systems and for process control. Capacitive measurement systems are often proposed for this application. In this method, electrodes are placed on the conveyor systems transport line and capacitive signals are sensed. The design of the sensor with regard to the arrangement and the number of electrodes as well as the evaluation of the capacitive sensor signals can be divided into two categories. Calibration-based flow meters use regression methods for signal processing, which are parametrized from calibration measurements on test rigs. Their performance is limited by the extend of the calibration measurements. Electrical capacitance tomography based flow meters use model-based signal processing techniques to obtain estimates about the spatial material distribution within the sensor. In contrast to their calibration-based counterparts, this approach requires more effort with respect to modeling and instrumentation, as typically a larger number of measurement signals has to be acquired. In this work we present a comparative analysis of the two approaches, which is based on measurement experiments and a holistic system model for flow metering. For the model-based analysis Monte Carlo simulations are conducted, where randomly generated pneumatic conveying flow patterns are simulated to analyze the sensor and algorithm behavior. The results demonstrate the potential benefit of electrical capacitance tomography based flow meters over a calibration-based instrument design.

Model-based signal processing enables bidirectional inferring between local field potential and spikes evoked by noxious stimulation

BackgroundRecording spontaneous and evoked activities by means of unitary extracellular recordings and local field potential (LFP) are key understanding the mechanisms of neural coding. The LFP is one of the most popular and easy methods to measure the activity of a population of neurons. LFP is also a composite signal known to be difficult to interpret and model. There is a growing need to highlight the relationship between spiking activity and LFP. Here, we hypothesized that LFP could be inferred from spikes under evoked noxious conditions. MethodRecording was performed from the medullary dorsal horn (MDH) in deeply anesthetized rats. We detail a process to highlight the C-fiber (nociceptive) evoked activity, by removing the A-fiber evoked activity using a model-based approach. Then, we applied the convolution kernel theory and optimization algorithms to infer the C-fiber LFP from the single cell spikes. Finally, we used a probability density function and an optimization algorithm to infer the spikes distribution from the LFP. ResultsWe successfully extracted C-fiber LFP in all data recordings. We observed that C-fibers spikes preceded the C-fiber LFP and were rather correlated to the LFP derivative. Finally, we inferred LFP from spikes with excellent correlation coefficient (r = 0.9) and reverse generated the spikes distribution from LFP with good correlation coefficients (r = 0.7) on spikes number. ConclusionWe introduced the kernel convolution theory to successfully infer the LFP from spikes, and we demonstrated that we could generate the spikes distribution from the LFP.

Design for parallel computation of model-based signal processing in Thomson scattering diagnostic

The Thomson Scattering (TS) diagnostic system calculates electron temperature and density profiles of tokamak plasma. The diagnostic is enabled by measuring and analyzing pulse signals of scattered laser light in the tokamak plasma. In order to improve accuracy of the diagnostic, it is necessary to reduce noise level of the pulse signals. In KSTAR, a fast digitizer with 5 GS/s has been adopted for obtaining the pulse shape information, and a research of model-based signal processing for the pulse signals has been conducted. In this paper, we present a new approach for the model-based signal processing and its parallel computation architecture design to improve its computation speed. As our approach, we conduct a nonlinear least square method in frequency domain by taking Fourier Transformation (FT) of the TS pulse model containing a convolution operator. Then, we can obtain algebraic functions for the transformed model and its Jacobian, which can reduce amount of computation and can be computed in parallel respect to frequency component. We implement the parallel computation of the algorithm in Graphical Processing Unit (GPU) and evaluate its effectiveness in terms of computation performance, convergence analysis, and variation of the model parameters.

Marine Icing Sensor with Phase Discrimination.

In this paper, a novel approach is presented to the measurement of marine icing phenomena under the presence of a two-phase condition. We have developed a sensor consisting of an electrostatic array and a signal processing based on a decision tree method. A three-element electrostatic array is employed to derive signals having linearly decoupled characteristics from which two key parameters, ice and water accretion layer dimension, can be determined for the purpose of environmental monitoring. The quantified characteristics revealed a correlation with the ice layer thickness in spite of the strong influence from the top water phase layer. The decision tree model established a relationship between the signal characteristics and the two accretion thickness parameters of water and ice layer. Through experimental verification, it has been observed that our sensor array in combination with the decision tree model based signal processing provides a simple practical solution to the challenging field of a two phase composition measurement such as in the marine icing considered in this study.

Gear Shape Parameter Measurement Using a Model-Based Scanning Multi-Distance Measurement Approach.

To reduce wind turbine failures by defective drive trains, deviations in the geometry of large gears (diameter ≳ 1 m) must be extensively determined with single-digit micrometer uncertainties. Fixed measuring volumes limit standard measuring methods like coordinate and gear measuring instruments for large gear measurements. Therefore, a model-based scanning multi-distance measurement approach for gear shape parameters is presented. The measurement approach has a scalable design and consists of a confocal-chromatic sensor, rotary table as a scanning unit and model-based signal processing. A preliminary study on a midsize spur gear demonstrates the general feasibility of the model-based scanning multi-distance measurement approach. As a result, the mean base circle radius as the fundamental gear shape parameter is determined with an uncertainty of <5 μm. The calibration and adjustment of the sensor arrangement were performed with a known calibration gear. Scalability is not experimentally validated in this article. However, simulations verify the scalability of the measurement approach in a first step. For gears with 1 m in diameter and varying tooth flank geometries, the estimated achievable uncertainty of the mean base circle radius is still <5 μm. Therefore, the model-based scanning multi-distance measurement approach is a promising alternative for gear inspection.

Indirect fluorescence-based in situ geometry measurement for laser chemical machining

Optical geometry measurements of submerged workpiece surfaces in the chaotic fluid environment of laser chemical machining (LCM) are challenging, because the produced cavities feature high aspect ratios and surface gradients. To avoid reflection-based artefacts at high gradients, molecules are added to the acid, whose fluorescence is detected with a confocal setup. The model-based signal processing enables the indirect measurement of the micro-geometry even in acid layers with mm to cm thickness and is capable to cope with process-inherent bubbles. As a result, the geometry measurement is shown to be applicable for the LCM process at surface gradients up to 84°.

Mitigation of RF Impairments of a 160-GHz MMIC FMCW Radar Using Model-Based Estimation

Employing frequency bands above 100 GHz for future frequency-modulated continuous-wave (FMCW) radar applications necessitate hardware realizations as monolithic microwave integrated circuit (MMIC). Spurious signals stemming from hardware impairments deteriorate target detection performance, but hardware-based mitigation is not preferred, given the increased cost and size of the integrated circuits. Instead, signal processing-based mitigation is favored. In this article, an approach is proposed to mitigate hardware impairments by parametric, model-based signal processing. For one particular FMCW radar operating at 160 GHz, a behavioral model of the radar device is developed, which accounts for the hardware impairments. This device model is incorporated in the data model of a maximum-likelihood parameter estimator that both resolves target ranges and mitigates spurious signal components. The mitigation performance and the improved robustness of target detection of this approach in the presence of hardware impairments are demonstrated by measurements.

Ed Sullivan: Mentor, scientist, colleague, and friend

I had the great fortune of meeting Ed Sullivan in 1989 when he was the Manager of Basic Research at the Naval Undersea Warfare Center and, to his dismay, was spending almost all of his time managing rather than researching. Ed hired me as a part-time student associate to analyze and implement passive synthetic aperture sonar, the core of our collaborative efforts for over twenty years and the technical focus of this presentation. Conversations with Ed, whether technical or not, were intellectually stimulating, fun, or mostly a combination of the two. Importantly though and through Ed’s confident and unique lens, he provided me – and many of you – with insights into highly relevant areas such as matched field and model based signal processing, music, history, and physics!I had the great fortune of meeting Ed Sullivan in 1989 when he was the Manager of Basic Research at the Naval Undersea Warfare Center and, to his dismay, was spending almost all of his time managing rather than researching. Ed hired me as a part-time student associate to analyze and implement passive synthetic aperture sonar, the core of our collaborative efforts for over twenty years and the technical focus of this presentation. Conversations with Ed, whether technical or not, were intellectually stimulating, fun, or mostly a combination of the two. Importantly though and through Ed’s confident and unique lens, he provided me – and many of you – with insights into highly relevant areas such as matched field and model based signal processing, music, history, and physics!

A few comments on passive synthetic aperture sonar in honor of Ed Sullivan

I came to know Ed Sullivan after submitting a paper on passive synthetic aperture sonar for publication in the Journal of the Acoustical Society of America. At that time, he was an Associate Editor of the journal. I knew of him, by reputation, of course. Only after the paper reviews came back did I get to know him, the person, the hard-core, old-school signal processor and physicist. At first, I simply tried to humor him to get the paper accepted for publication, figuring that quotes from a few of his papers sprinkled here and there would suffice. But after spending time expanding my knowledge of his work, I developed a deep appreciation for the foundation in signal processing he and colleagues had built. The published version of our paper was significantly improved as a result. This talk presents some of Ed Sullivan's results on model-based signal processing and its application to passive synthetic aperture sonar, along with some of our published results. Its main purpose, however, is to serve as a reminder that we all stand on the foundation built by those who have gone before. [Work supported by the Office of Naval Research.]

Ed Sullivan and model-based ocean acoustic signal processing

Ed Sullivan was a pioneer in model-based signal processing and his contributions have inspired numerous scientists in the fields of acoustic signal processing and ocean acoustics. Work by Ed and his colleagues is among the first efforts to formulate ocean acoustic inverse problems in a sequential filtering framework; methods were developed that overcome the destructive role of mismatch between assumed and true physics in inverse problems. In paper “Model-based ocean acoustic signal processing,” Acoustics Today, July 2011, Ed Sullivan identifies three main directions in signal processing, detection, classification, and estimation; he focuses on estimation and presents a study of model-based methods for inverse problems and how those methods can be successfully applied to the field of ocean acoustics. This presentation reviews these applications and discusses the impact of Ed Sullivan’s contributions on other advances in ocean acoustic signal processing.

Multilayer Thickness Measurements below the Rayleigh Limit Using FMCW Millimeter and Terahertz Waves

We present thickness measurements with millimeter and terahertz waves using frequency-modulated continuous-wave (FMCW) sensors. In contrast to terahertz time-domain spectroscopy (TDS), our FMCW systems provide a higher penetration depth and measurement rates of several kilohertz at frequency modulation bandwidths of up to 175 GHz. In order to resolve thicknesses below the Rayleigh resolution limit given by the modulation bandwidth, we employed a model-based signal processing technique. Within this contribution, we analyzed the influence of multiple reflections adapting a modified transfer matrix method. Based on a brute force optimization, we processed the models and compared them with the measured signal in parallel on a graphics processing unit, which allows fast calculations in less than 1 s. TDS measurements were used for the validation of our results on industrial samples. Finally, we present results obtained with reduced frequency modulation bandwidths, opening the window to future miniaturization based on monolithic microwave integrated circuit (MMIC) radar units.

A New Model Based Signal Processing Scheme for Detecting Delay and Doppler in Bistatic Radars

In this paper, a new method is presented for joint estimation of delay and Doppler in bistatic radars by using model-based signal processing framework. In the proposed method, the time history of each cell in ambiguity function is modeled as a stochastic process consisting of a stationary component caused by clutter and noise as well as a possible transient component which is caused by a target. Then some cells are indicated as candidates for being targeted based on estimating the higher order statistics of the above stochastic processes. Finally, the spatial processing scheme is performed to prune false candidates. To evaluate the proposed method, its performance is simulated under two different scenarios (i.e., slow and fast moving targets) which have been inspirited from real conditions. Comparison of the results obtained from the proposed method with its alternatives, shows its better performance in such a way that it detects fast targets at least 8% better than the best among all examined methods. Furthermore, the proposed method detects slow targets at least 4.3% better than the best among all examined methods. Both of the above superiorities are obtained in those conditions while the false alarm of the proposed method does not show the meaningful difference against other examined algorithms.

Signal processing techniques for rolling element bearing spall size estimation

Abstract This paper proposes signal processing techniques to extract entry and exit points from the bearing vibration signal for spall size estimation. The entry point is the start point of a low frequency response when a rolling element enters a localized defect on the raceway, which is contaminated with background noises and difficult to identify. An empirical model based signal processing method is proposed to effectively identify the entry point. The Variational Mode Decomposition (VMD) is applied to the bearing entry signal for more accurate estimation. Differentiation technique is used to identify the high frequency exit point with more reliable threshold values for automatic diagnostics. Then, based on defect size estimation models for both inner and outer race defects, the spall size can be estimated. The proposed methodology is validated on a machine tool spindle’s bearing system. Experimental results show that the proposed signal processing techniques provide less biased results with respect to spindle speed and more precise estimation.

Model-Based Position and Reflectivity Estimation of Fiber Bragg Grating Sensor Arrays.

We propose an efficient model-based signal processing approach for optical fiber sensing with fiber Bragg grating (FBG) arrays. A position estimation based on an estimation of distribution algorithm (EDA) and a reflectivity estimation method using a parametric transfer matrix model (TMM) are outlined in detail. The estimation algorithms are evaluated with Monte Carlo simulations and measurement data from an incoherent optical frequency domain reflectometer (iOFDR). The model-based approach outperforms conventional Fourier transform processing, especially near the spatial resolution limit, saving electrical bandwidth and measurement time. The models provide great flexibility and can be easily expanded in complexity to meet different topologies and to include prior knowledge of the sensors. Systematic errors due to crosstalk between gratings caused by multiple reflections and spectral shadowing could be further considered with the TMM to improve the performance of large-scale FBG array sensor systems.

Model-Based Signal Processing for the Force Control of Biaxial Gantry Robots

For the automation of certain industrial tasks, like the handling of deformable materials, force/torque sensors are frequently employed in robotic manipulators. This work investigates different signal processing strategies for a force sensor mounted on the end-effector of a biaxial gantry robot. For this, a suitable mathematical model of the manipulator is derived and analyzed. The merits and limitations of the proposed signal processing strategies are compared and validated by a number of experimental results.

Target localization in a reverberant shallow ocean waveguide with environmental uncertainty using a nonlinear frequency-difference signal processing technique

Model-based signal processing for active sonar localization is typically infeasible due to insufficient knowledge of the acoustic environment. Additionally, in a shallow ocean environment, surface or bottom roughness create diffuse reverberation that can often obscure a desired target echo. Recently, a nonlinear signal processing technique was developed for passive acoustic source localization (Worthmann, et al., 2015) that is applicable to high frequency sources in uncertain shallow ocean environments. This technique exploits a nonlinear field product (the autoproduct) and uses in-band hydrophone array measurements to determine field information in a lower, out-of-band, frequency regime where environmental uncertainties are less detrimental. When extended to monostatic active sonar with a vertical array, this technique allows a model-based signal processing algorithm to combat the detrimental effects of reverberation. The nonlinear signal processing algorithm is presented, along with simulations in a 5-km range, 200-m deep ideal waveguide with environmental uncertainties and significant reverberation at frequencies between 2- and 5-kHz. Successful detection and localization of a mid-water-column target is found to be possible at simulated signal-to-reverberation levels as low as −5 dB. Comparisons to existing signal processing detection and localization algorithms are provided. [Sponsored by the Office of Naval Research and the National Science Foundation.]

Comparison of Adaptive and Model-Free Methods for Dynamic Measurement

Dynamic measurement aims to improve the speed and accuracy characteristics of measurement devices by signal processing. State-of-the-art dynamic measurement methods are model-based adaptive methods, i.e., 1) they estimate model parameters in real-time and 2) based on the identified model perform model-based signal processing. The proposed model-free method belongs to the class of the subspace identification methods. It computes directly the quantity of interest without an explicit parameter estimation. This allows efficient computation as well as applicability to general high order multivariable processes.

Fetal QRS extraction from abdominal recordings via model-based signal processing and intelligent signal merging

Noninvasive fetal ECG (fECG) monitoring has potential applications in diagnosing congenital heart diseases in a timely manner and assisting clinicians to make more appropriate decisions during labor. However, despite advances in signal processing and machine learning techniques, the analysis of fECG signals has still remained in its preliminary stages. In this work, we describe an algorithm to automatically locate QRS complexes in noninvasive fECG signals obtained from a set of four electrodes placed on the mother’s abdomen. The algorithm is based on an iterative decomposition of the maternal and fetal subspaces and filtering of the maternal ECG (mECG) components from the fECG recordings. Once the maternal components are removed, a novel merging technique is applied to merge the signals and detect the fetal QRS (fQRS) complexes. The algorithm was trained and tested on the fECG datasets provided by the PhysioNet/CinC challenge 2013. The final results indicate that the algorithm is able to detect fetal peaks for a variety of signals with different morphologies and strength levels encountered in clinical practice.