Browsing by Subject "Signal processing"

Now showing 1 - 20 of 27

Admittance derived stroke volume for determination of hemodynamic stability during atrial and ventricular arrhythmias
(2017-12) Holt, Lucas; Valvano, Jonathan W., 1953-; Feldman, Marc D.; Pearce, John A.; Rylander, Henry G.; Swartzlander, Earl E.
Implantable cardioverter defibrillators (ICDs) are medical devices proven to prevent sudden cardiac death due to ventricular arrhythmias. Their decisions are based upon intra-cardiac electrograms (IEGM). This is incomplete information since up to 5% of implantable cardioverter defibrillator (ICD) shocks are inappropriate. Receiving a shock is associated with increased mortality as well as emotional trauma. In contrast, physicians determine whether to shock a patient out of a rapid rhythm by determining if the arrhythmia is hemodynamically unstable or stable. An unstable arrhythmia is identified by decreased forward stroke volume (SV) and resultant low blood pressure (BP). It would be ideal to have beat-by-beat SV available to the ICD to assist in the delivery of therapies. A system that utilizes the right ventricular (RV) shocking lead of an ICD to measure the electrical admittance in the RV is proposed for measuring continuous SV. For this method to work, a signal processing technique to remove noise artifacts related to lead motion and respiration must be developed.
Analytics and their applications to power quality and power system data
(2020-06-22) Furlani Bastos, Alvaro; Santoso, Surya; Baldick, Ross; Vikalo, Haris; Zhu, Hao; Freitas, Walmir
Multiple monitoring devices deployed throughout transmission and distribution networks enable insightful analysis of power systems' behavior. These devices record signals either at the waveform or phasor levels, and their time resolution range from milliseconds to minutes. This work focuses primarily on the analysis and applications of power quality data at the waveform level. Also, potential applications based on other sources of power system data (namely, smart meters and phasor measurement units) are proposed as a minor contribution of this work. Modern power quality monitors have the capability of continuously storing the measured waveforms, unlike older devices where only a few data samples were recorded once a disturbance was detected. This evolution created the possibility of new and advanced applications of power quality data analytics. Whereas triggered power quality data have been used only for troubleshooting in the past, triggerless measurements contain more valuable information about the system, which allow event root causes identification and monitoring of equipment health and performance. However, analysis of triggerless power quality data requires parsing through large datasets for finding information hidden in the raw data. In this context, this work proposes two approaches for identifying disturbances and anomalous behavior on triggerless power quality data, which are based either on voltage/current waveforms or rms profiles. The waveform-based detector is a general framework for identifying shape changes between successive cycles of data. On the other hand, rms-based detectors are more suitable for applications related to switching operations, which create a step change in rms voltage/current profiles (such as capacitor switching and voltage regulator operation). In addition, two power quality data analytics applications are presented. First, a novel method is proposed for accurately characterizing voltage variation events (sags and swells), focused on determining their exact point-on-wave inception and recovery instants. Subsequently, triggered power quality data are employed for condition monitoring of circuit switchers for shunt capacitor banks at a transmission network; this analysis evaluates the performance of pre-insertion resistor/inductor and their switchgear during energizing operations. Finally, data analytics applications related to alternative power system data sources are discussed. Initially, machine learning techniques are used for predicting in advance voltage magnitudes throughout a distribution network, so that voltage regulation control is enhanced. Next, field data from phasor measurement units are used for illustrating some of the challenges encountered in system frequency estimation; then, a robust frequency estimation technique based on numerical derivatives is proposed for overcoming those limitations.
Computational process networks : a model and framework for high-throughput signal processing
(2011-05) Allen, Gregory Eugene; Evans, Brian L. (Brian Lawrence), 1965-; Browne, James C.; Chase, Craig M.; John, Lizy K.; Loeffler, Charles M.
Many signal and image processing systems for high-throughput, high-performance applications require concurrent implementations in order to realize desired performance. Developing software for concurrent systems is widely acknowledged to be difficult, with common industry practice leaving the burden of preventing concurrency problems on the programmer. The Kahn Process Network model provides the mathematically provable property of determinism of a program result regardless of the execution order of its processes, including concurrent execution. This model is also natural for describing streams of data samples in a signal processing system, where processes transform streams from one data type to another. However, a Kahn Process Network may require infinite memory to execute. I present the dynamic distributed deadlock detection and resolution (D4R) algorithm, which permits execution of Process Networks in bounded memory if it is possible. It detects local deadlocks in a Process Network, determines whether the deadlock can be resolved and, if so, identifies the process that must take action to resolve the deadlock. I propose the Computational Process Network (CPN) model which is based on the formalisms of Kahn’s PN model, but with enhancements that are designed to make it efficiently implementable. These enhancements include multi-token transactions to reduce execution overhead, multi-channel queues for multi-dimensional synchronous data, zero-copy semantics, and consumer and producer firing thresholds for queues. Firing thresholds enable memoryless computation of sliding window algorithms, which are common in signal processing systems. I show that the Computational Process Network model preserves the formal properties of Process Networks, while reducing the operations required to implement sliding window algorithms on continuous streams of data. I also present a high-throughput software framework that implements the Computational Process Network model using C++, and which maps naturally onto distributed targets. This framework uses POSIX threads, and can exploit parallelism in both multi-core and distributed systems. Finally, I present case studies to exercise this framework and demonstrate its performance and utility. The final case study is a three-dimensional circular convolution sonar beamformer and replica correlator, which demonstrates the high throughput and scalability of a real-time signal processing algorithm using the CPN model and framework.
Deep downhole testing: procedures and analysis for high-resolution vertical seismic profiling
(2008-05) Li, Songcheng, 1968-; Stokoe, Kenneth H.
A study was undertaken to improve the signal quality and the resolution of the velocity profile for deep downhole seismic testing. Deep downhole testing is defined in this research as measurements below 225 m (750 ft). The study demonstrated that current testing procedures can be improved to result in higher signal quality by customizing the excitation frequency of the vibrator to local site conditions of the vibrator-earth system. The earth condition beneath the base plate can be an important factor in the signal quality subject to variations with time when tests are repetitive. This work proposes a convenient method to measure the site localized natural frequency and damping ratio, and recommends using different excitation frequencies for P- and S-wave generation. Properly increasing the excitation duration of the source signal also contributes to the quality of the receiver signal. The source signature of sinusoidal vibratory source is identified. Conventional travel time analysis using vibratory source generally focuses on chirp sweeps. After testing with impulsive sources and chirp sweeps and comparing the results with the durational sinusoidal source, the sinusoidal source was then chosen. This work develops an approach to identifying the source signature of the sinusoidal source and concludes that the normalized source signature is relevant only to four parameters: the fixed-sine excitation frequency, the duration of excitation, the damping ratio of the vibrator-earth system, and the damped natural frequency of the vibrator-earth system. Two of the parameters are designated input to the vibrator and the other two parameters are measured in the field test using the proposed method in this work. A new wavelet-response technique based on deconvolution and consideration of velocity dispersion is explored in travel-time analyses. The wavelet-response technique is also used for development of a new approach to correcting disorientation of receiver tool. The improved downhole procedures and analyses are then used in the analysis of deep downhole test data obtained at Hanford, WA. Downhole testing was performed to a depth of about 420 m (1400 ft) at Hanford site. Improvements in resolving the wave velocity profiles to depths below 300 m (1000) ft are clearly shown.
Design and implementation of an underwater acoustic transponder
(2011-05) Perrine, Kenneth Avery; Evans, Brian L. (Brian Lawrence), 1965-; Hall, Neal A.
A transponder for underwater acoustic data communications is prototyped. The mobile transponder emits a data sequence whenever it detects a ping from a base station. The data sequence includes GPS coordinates and UTC time sent over a conservative and brief 12 kbps turbo-coded BPSK link, and a 6 kB JPEG image sent over an ambitious 67 kbps turbo-coded 16-QAM link. The range of the transponder from the base station can also be accurately derived. Several challenges exist in decoding the underwater signals at the base station receiver, including Doppler distortion and multipath. While experimental results show that the ranges for decoding the 16-QAM signals with a single hydrophone are limited to less than 25 m, the BPSK signals prove to be much more robust, decoding at ranges of up to 625 m. Experiments with delays and transducer tether length indicate methods for improving reliability in the presence of reverberation and thermocline. This transponder uses mostly off-the-shelf parts and is anticipated to be improved when paired with advanced sonar array devices.
Diffraction imaging by path-summation migration
(2018-08-07) Merzlikin, Dmitrii; Fomel, Sergey B.; Foster, Douglas J; Ghattas, Omar; Meckel, Timothy A; Sen, Mrinal K
Unconventional reservoir characterization requires accurate and high-resolution subsurface images to detect small-scale geological features controlling the production efficiency. Diffraction imaging techniques provide higher lateral resolution images in comparison with the results of conventional reflection imaging and highlight direct responses of such subsurface discontinuities as faults, channel edges, fracture swarms and pinch-outs, distribution of which can be crucial for reservoir development decisions. There are three major challenges in diffraction imaging: reflection/diffraction separation, imaging of diffractions, and de-noising of diffractions. I develop a diffraction imaging workflow based on least-squares inversion and path-summation migration to address these challenges and to improve the robustness of diffraction imaging. Conventional reflection images are dominated by high-energy reflections, which mask diffractions. The challenge of diffraction/reflection separation is to extract diffracted energy by suppressing reflections. Diffraction on an edge has both reflective and diffractive components. Both components should be preserved to generate a diffraction image. I develop azimuthal plane-wave destruction workflow (AzPWD) to account for edge diffraction signatures. The method suppresses high-energy reflections and preserves edge diffractions by orienting plane wave destruction (PWD) filter perpendicular to the edge. Edge orientations are also determined and can be utilized for the interpretation. Diffraction imaging is based on conventional imaging operators tailored towards reflections. I develop analytical expressions for path-summation integral diffraction imaging, which naturally incorporates diffraction apex stationarity under time migration velocity perturbation. This approach allows for determining diffraction likelihood distribution with the approximate cost of only two fast Fourier transforms in a velocity-model-independent fashion. I also develop double-path-summation framework for automatic migration velocity analysis based on diffractions, which does not require picking. Diffraction images are prone to noise and contain reflection remainders after application of reflection-diffraction separation procedure. I address this problem using least-squares migration. I define the inverted forward modeling operator as the chain of three operators: Kirchhoff modeling, plane-wave destruction and path-summation integral filter. This chain of operators accounts for diffraction energy contribution to the least-squares migration misfit dominated by reflections. I propose to use sparsity constraints to penalize diffractions with spiky and intermittent distributions. Reflections are regularized using smoothing along the dominant local slopes in the image domain. I use a shaping regularization framework. The approach decomposes input data into diffractions, reflections and noise. I extend the proposed chain inversion approach to 3D to account for edge diffraction responses by replacing PWD with AzPWD in forward modeling. I penalize edge diffractions using both sparsity constraints and anisotropic diffusion. The first regularization extracts diffractive component of the edge response whereas the second one enforces continuity along the edge to account for the reflective component. Proposed inversion schemes address the challenge of diffraction de-noising and can be treated as imaging operators tailored towards diffractions and extracting edge diffractions in an iterative fashion. The developed workflows allow for diffraction extraction from reflections and noise and for accurate focusing of diffracted energy. Numerous synthetic and field data examples are used to test the performance of the proposed methods. The tests confirm their effectiveness.
Efficient channel estimation for block transmission systems
(2006) Shin, Changyong; Powers, Edward J.
Block transmission systems have recently gained considerable interest as a promising method for high data rate communications. This is due to their uncomplicated implementation and simple equalization of frequency-selective fading channels. For coherent signal detection and channel equalization in block transmission systems, channel state information (CSI) should be known to, or estimated at, the receiver. In this dissertation, we present three approaches for efficient channel estimation in block transmission systems. First, to provide a bandwidth-efficient solution for multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) channel estimation, we establish conditions for channel identifiability and propose a blind channel estimation method based on a subspace technique. Second, to relax existing strict conditions for blind MIMO channel identification without a sacrifice of data rates and to provide a bandwidth-efficient solution for channel estimation in MIMO block transmission systems with a cyclic prefix, we present a framework for blind channel estimation based on a general non-redundant precoding. Using this framework, we propose a blind channel estimator exploiting a simplified non-redundant precoding. To complete the channel estimation, we also develop a technique for resolving the channel ambiguity in the proposed method. Third, in rapid mobile environments where channels change very fast, blind channel estimation techniques may not be suitable to obtain CSI due to their relatively slow convergence. In this case, to achieve accurate estimation of doubly selective channels in OFDM systems, we propose an optimal (in the sense of mean square error) pilot tone placement applicable to OFDM systems regardless of the time variations of a channel. In addition, we present an accurate linear minimum mean square error (LMMSE) channel estimator that exploits a small number of pilot tones located according to the derived optimal placement. To achieve computationally efficient channel estimation with lower complexity than the LMMSE estimator and to obtain performance close to the LMMSE estimator, an approximate LMMSE (ALMMSE) channel estimator is also proposed. Finally, we propose a novel iterative ALMMSE channel estimator that achieves better performance than the LMMSE and ALMMSE estimators, while having complexity in between the two.
Efficient signal acquisition and deep learning model compression
(2022-11-17) Sakthi, Madhumitha; Tewfik, Ahmed; Vikalo, Haris; Miikkulainen, Risto; Millan, Jose del R.; Hamilton, Liberty
The ubiquitous use of deep learning models for signal processing has led to an increasing computational and storage cost, especially in edge applications. Although deep learning has delivered improved accuracy for a given task, they have also increased the need for bringing in efficiency during signal processing and acquisition to save on computation, storage and overall power consumption. The main motivation of this thesis is to present algorithms for efficient signal acquisition and compress the deep learning models across various applications and use deep learning models to again increase processing efficiency. Typically, to make a decision on a given signal, it is vital to acquire robust signal information and use a processing chain that is capable of making a decision on the input with minimum resource utilization and present the output to the end user. In order to make this process efficient, we present our solutions in three parts. First, in autonomous driving applications where robust signal acquisition is essential for safety during deployment, we design a novel radar sub-sampling algorithm designed to pick regions of interest that need more accurate reconstruction thereby providing optimal performance at a considerably low sampling rate. We developed this method using images, images and radar, only radar for good and extreme weather conditions respectively. The algorithm is further improved to use the predicted motion of the object for region determination and we analyzed a hardware-efficient Binary Permuted Diagonal measurement matrix in compressed sensing and show competitive performance to the Gaussian measurement matrix. Finally, we trained a novel transformer-based object detection system for image and radar object detection and a state-of-the-art object detection system from only radar data. Second, for signal processing efficiency, we present our novel deep learning model compression algorithm in order to compress the deep learning model either for storage or storage and computational efficiency. The algorithms were designed for two conditions, retraining and non-retraining cases. Our Bin & Quant algorithm is thoroughly designed for compressing float models without retraining for a negligible loss in accuracy. The Bin & Quant algorithm is also applied directly to the integer quantized models and achieves storage compression on the already compressed and computationally efficient model. This method is specifically designed for compressing off-the-shelf models without the knowledge of hyperparameters or the availability of the original training data. Finally, for utmost compression, we present our Gradient Weighted K-means algorithm, designed to train computationally efficient integer quantized models where each weight requires less than 1-bit for storage. Using the above compression methods, we compressed the speech recognition and vision models. Third, we developed lightweight deep learning models for various BCI applications. Our main focus was to develop small networks for various applications in order to accommodate both computational and storage efficiency needs while also providing robust results. Notably, our keyword spotter is designed to invoke a larger EEG-based recognition system only when the user needs and therefore, switch on the larger recognition system on a need basis and a native language classification model can be used to direct EEG data to separate networks that process native EEG and non-native EEG-based speech recognition, thereby decreasing the complexity of the larger network. Similarly, the audio vs. audio-visual stimuli classification based on EEG is again intended for directing the EEG data to separate neural networks solely designed for either audio stimuli-based EEG or audio-visual stimuli-based EEG networks. Therefore, using the above three contributions, we aim at improving signal acquisition and processing efficiency using the deep learning models and also propose a novel deep learning model compression algorithm.
Electroencephalography (EEG) based speech technologies using neural networks
(2021-11-02) Krishna, Gautam, Ph. D.; Tewfik, Ahmed; Millan, Jose Del; Fernandez, Benito R; Julien, Christine L; Vikalo , Haris
The emergence of virtual personal assistants like Apple Siri, Amazon Alexa, Google Assistant, Samsung Bixby, Windows Cortana, etc has improved the user experience for smartphone and personal computer users. The automatic speech recognition (ASR) system forms an important component of a virtual assistant. An ASR system converts speech signals into text which is further processed by the natural language understanding (NLU) component of the virtual assistant. However, the performance of an ASR system degrades in presence of background noise and this affects the performance of virtual assistants in noisy environments like shopping mall or airport. On the other hand, studies have demonstrated that we humans are able to perform speech recognition with a lower word error rate (WER) compared to machines in presence of background noise. This motivated me to investigate how to use non-invasive electroencephalography (EEG) brain signal features recorded synchronously with noisy speech to improve the performance of ASR and other speech processing models. Current state-of-the-art ASR systems are trained to recognize only acoustic features and this limits technology accessibility for people with speaking disabilities. This motivated me to investigate techniques to design ASR systems capable of recognizing EEG features with no speech input. We, humans, have a high speech rate of 150 words per minute, thus an EEG speech prosthetic that works by first recognizing text from EEG signal and then generating speech from the recognized text using a state-of-the-art speaker dependent or independent text-to-speech (TTS) system, may suffer high latency. This motivated me to investigate algorithms to generate speech signals directly from EEG signals instead of translating EEG signals to text. First, In this thesis, we demonstrate a neural network-based algorithm to improve the performance of ASR and voice activity detection (VAD) systems operating in presence of background noise on a limited English vocabulary using EEG features. We also show that EEG features can be used to improve end-pointer detection model performance as an extension of VAD application. Second, In this thesis we demonstrate a neural network-based algorithm to perform isolated speech recognition using only EEG features with no speech with high accuracy on a limited English vocabulary and we then study three techniques inspired by representation learning to improve the performance of continuous speech recognition systems using only EEG features with no speech. Third, In this thesis, we study different techniques to generate speech features from EEG features and vice-versa. We study a recurrent neural network (RNN) based regression model with and without attention layer to generate acoustic features from EEG features. We demonstrate generating acoustic features from EEG features with low test time error rates using the RNN model with and without attention layer. We show that for majority of our experiments attention model outperformed the RNN model without attention layer in terms of test time error rates even though for some subjects adding attention layer to the model was not helpful. We further identified the right sampling frequency and acoustic feature dimension to generate audio waveform with broader characteristics closer to the ground truth audio waveform from EEG signals. Fourth, In this thesis, we demonstrate an algorithm to improve the speech recognition performance for aphasia, apraxia and dysarthria speech by utilizing EEG features. We demonstrate that our proposed algorithm in real-time can make use of ear EEG, dry EEG and acoustic features to outperform a baseline speech recognition model trained using only acoustic features when tested on several subjects with various severity levels of aphasia, apraxia and dysarthria. We further show that the algorithm can be extended to other tasks like speaker identification and voice activity detection for impaired speech. We demonstrate that the proposed algorithm can also be used to improve the performance of speech recognition systems operating in presence of background noise in addition to impaired speech.
Extraction of blade-vortex interactions from helicopter transient maneuvering noise
(2014-05) Stephenson, James Harold; Tinney, Charles Edmund, 1975-
Time-frequency analysis techniques are proposed as a necessary tool for the analysis of acoustics generated by helicopter transient maneuvering flight. Such techniques are necessary as the acoustic signals related to transient maneuvers are inherently unsteady. The wavelet transform is proposed as an appropriate tool, and it is compared to the more standard short-time Fourier transform technique through an investigation using several appropriately sized interrogation windows. It is shown that the wavelet transform provides a consistent spectral representation, regardless of employed window size. The short-time Fourier transform, however, provides spectral amplitudes that are highly dependent on the size of the interrogation window, and so is not an appropriate tool for this situation. An extraction method is also proposed to investigate blade-vortex interaction noise emitted during helicopter transient maneuvering flight. The extraction method allows for the investigation of blade-vortex interactions independent of other sound sources. The method is based on filtering the spectral data calculated through the wavelet transform technique. The filter identifies blade-vortex interactions through their high amplitude, high frequency impulsive content. The filtered wavelet coefficients are then inverse transformed to create a pressure signature solely related to blade-vortex interactions. This extraction technique, along with a prescribed wake model, is applied to experimental data extracted from three separate flight maneuvers performed by a Bell 430 helicopter. The maneuvers investigated include a steady level flight, fast- and medium-speed advancing side roll maneuvers. A sensitivity analysis is performed in order to determine the optimal tuning parameters employed by the filtering technique. For the cases studied, the optimized tuning parameters were shown to be frequencies above 7 main rotor harmonics, and amplitudes stronger than 25% (−6 dB) of the energy in the main rotor harmonic. Further, it is shown that blade-vortex interactions can be accurately extracted so long as the blade-vortex interaction peak energy signal is greater or equal to the energy in the main rotor harmonic. An in-depth investigation of the changes in the blade-vortex interaction signal during transient advancing side roll maneuvers is then conducted. It is shown that the sound pressure level related to blade-vortex interactions, shifts from the advancing side, to the retreating side of the vehicle during roll entry. This shift is predicted adequately by the prescribed wake model. However, the prescribed wake model is shown to be inadequate for the prediction of blade-vortex interaction miss distance, as it does not respond to the roll rate of the vehicle. It is further shown that the sound pressure levels are positively linked to the roll rate of the vehicle. Similar sound pressure level directivities and amplitudes can be seen when vehicle roll rates are comparable. The extraction method is shown to perform admirably throughout each maneuver. One limitation with the technique is identified, and a proposal to mitigate its effects is made. The limitation occurs when the main rotor harmonic energy drops below an arbitrary threshold. When this happens, a decreased spectral amplitude is required for filtering; which leads to the extraction of high frequency noise unrelated to blade-vortex interactions. It is shown, however, that this occurs only when there are no blade-vortex interactions present. Further, the resulting sound pressure level is identifiable as it is significantly less than the peak blade-vortex interaction sound pressure level. Thus the effects of this limitation are shown to be negligible.
High-performance [delta sigma] analog-to-digital conversion
(2008-05) Tsang, Robin Matthew, 1979-; Valvano, Jonathan W., 1953-
This dissertation is about a new [delta sigma] analog-to-digital converter that offers enhanced quantization noise suppression at low oversampling ratios. This feature makes the converter attractive in applications where speed and resolution are simultaneously demanded. The converter exploits double-sampling for speed, and takes advantage of a new loop-filter to pin down passband quantization noise. A proto-type is fabricated in 0.18-[mu]m CMOS and tested. Results show that at 200-MS/s, the converter achieves an effective number of bits (ENOB) of 12.2-b in a 12.5-MHz signal band while consuming 89-mW from a 1.8-V supply. Using a common performance metric that takes into account of ENOB and signal bandwidth, the prototype outperforms all previously-reported IEEE switched-capacitor [delta sigma] modulators.
The importance of sediment roughness on the reflection coefficient for normal incidence reflections
(2011-05) Hron, Joel Maurice; Isakson, Marcia J.; Ezekoye, Ofodike A.
This research experimentally shows the effect of sediment roughness characteristics on the acoustic reflection coefficient. This information is useful when trying to classify various types of sediment over an area. This research was conducted in an indoor laboratory tank at Applied Research Laboratories (ARL) at the University of Texas at Austin. A single beam echo-sounder (SBES) system was developed to project and receive a wideband (3 kHz to 30 kHz) acoustic pulse. A method was developed using the system transfer function to create a custom pulse that would minimize the dynamic range over the wide frequency band. A matched filtering and data processing algorithm was developed to analyze data over the full frequency bandwidth and over smaller frequency bands. Analysis over the smaller frequency bands showed the effect of the roughness on the reflection coefficient with respect to frequency. It was found that the reflection coefficient is significantly lower at the higher frequencies (above 20 kHz) than at the lower frequenices [sic] due to off specular scattering. It was also found that the variability of the reflection coefficient was significantly higher for the rough sediment than for the smooth sediment.
Incorporation of the Global Positioning System modernization signals into existing smoother-based ephemeris generation processes
(2008-05) Harris, Robert B., Ph. D.; Lightsey, E. Glenn
The introduction of M-Code to the GPS signal structure can redefine the accuracy of the broadcast ephemeris. Existing ephemeris generation systems use dual frequency observations, obtained through the tracking of existing precise codes on the L1 and L2 frequencies. These codes are modulated using Binary Phase Shift Key (BPSK) modulation. The modernization signal M-Code is modulated using Binary Offset Carrier (BOC) modulation. In this study pseudorange observables derived from the tracking of M-Code are proven to have greater accuracy than those from existing precise codes, given equivalent receiver designs and operating conditions. In addition, the error due to specular multipath is derived. These general models of noise and multipath can be applied to any BOC modulated signals, including Galileo and QZSS. When applied to M-Code, the models predict that the maximum multipath error in the pseudorange is reduced in magnitude by 50% compared to the existing precise codes. However the range of multipath delays for which M-Code observables exhibit multipath is approximately twice that associated with existing precise BPSK codes. Existing ephemeris generation processes use the ionosphere free combination and carrier phase smoothing of the pseudorange to form smoothed pseudoranges. The smoothed pseudoranges are then input as measurements to an ephemeris filter. The analytic models of multipath error in the pseudorange and carrier phase observables are applied to predict errors in the smoothed pseudorange. Multipath error, amplified by ionosphere free combination, causes a bias in the smoothed pseudorange when parameterized as a function of multipath delay. There are conditions under which the bias is zero mean, and in those conditions multipath is suppressed. The mechanism for those conditions is solved and discussed, for both BOC and BPSK signal tracking. The solution of carrier phase multipath for BOC modulated signals also admits solutions with a special quality not seen in the BPSK solution. There are multipath delays for which the carrier phase multipath is identically zero regardless of the multipath phase. The zero carrier phase multipath condition may be the most promising feature associated with observables derived from BOC modulated codes.
Linearity analysis of microwave photonic links for analog signal processing
(2022-07-01) Mokhtari Koushyar, Farzad; Bank, Seth Robert; Vishwanath, Sriram; Campbell, Joe C; Wasserman, Daniel M; Nanzer, Jeffrey
Microwave photonics (MWP) provides wideband, programmable, and low-loss platforms for analog signal processing with high power handling and immunity to electromagnetic interference. Recent advancements in integrated photonics have enabled a variety of on-chip functions for signal processing including true-time delays, tunable switches, high-Q resonators, frequency combs, and so on which make MWP a promising solution for ever-increasing demand of high throughput signal processing. However, achieving the desired dynamic range (DR) from MWP links has remained elusive for signal processing applications. The nonlinear sources of MWP links are studied with a focus on integrated MWP for signal processing. The concept of interference induced distortions (IIDs) are introduced which are generated by passive structures in the link. As demonstrated by the presented theory, simulations, and measurement results, IIDs from passive components can dominate the nonlinear distortions of active components in the link by tens of dB. The impact of parasitic interferometric structures which are formed by design, fabrication, and packaging imperfections on IIDs are discussed along with mitigation solutions. The impact of frequency chirp in active devices on IIDs are studied by simulation and measurement results which shows a magnified sensitivity of IIDs to parasitic interferences when chirp increases. On-chip tap combination, however, is a desired interference in many applications which its implementation remains challenging due to instabilities and limited DR originating from IIDs. Incoherent combiners are proposed in the literature based on multiple PDs or wavelengths at the cost of limiting bandwidth, increasing insertion loss, and necessitating a frequency comb. A tapered-pitch array combiner (TPAC) is introduced here to break the trade-offs in the incoherent and off-chip solutions. TPAC is designed based on the rules derived from the theory and simulation presented for modeling IIDs. A SFDR₃ equal to 107.3 dBc.Hz [superscript 2/3] is measured for the fabricated TPAC combing taps of a 6-tap filter using one wavelength and one PD. A modified TPAC is proposed for further IID suppression using optical phase alignment (OPA) of taps where up to than 26.9 dB IMD₃ suppression is measured. Finally, the stability and tunability of OPA approach are discussed followed by suggestions for future research.
Manifold signal processing for MIMO communications
(2009-12) Inoue, Takao, doctor of electrical and computer engineering; Heath, Robert W., Jr, 1973-
The coding and feedback inaccuracies of the channel state information (CSI) in limited feedback multiple-input multiple-output (MIMO) wireless systems can severely impact the achievable data rate and reliability. The CSI is mathematically represented as a Grassmann manifold or manifold of unitary matrices. These are non-Euclidean spaces with special constraints that makes efficient and high fidelity coding especially challenging. In addition, the CSI inaccuracies may occur due to digital representation, time variation, and delayed feedback of the CSI. To overcome these inaccuracies, the manifold structure of the CSI can be exploited. The objective of this dissertation is to develop a new signal processing techniques on the manifolds to harvest the benefits of MIMO wireless systems. First, this dissertation presents the Kerdock codebook design to represent the CSI on the Grassmann manifold. The CSI inaccuracy due to digital representation is addressed by the finite alphabet structure of the Kerdock codebook. In addition, systematic codebook construction is identified which reduces the resource requirement in MIMO wireless systems. Distance properties on the Grassmann manifold are derived showing the applicability of the Kerdock codebook to beam-forming and spatial multiplexing systems. Next, manifold-constrained algorithms to predict and encode the CSI with high fidelity are presented. Two prominent manifolds are considered; the Grassmann manifold and the manifold of unitary matrices. The Grassmann manifold is a class of manifold used to represent the CSI in MIMO wireless systems using specific transmission strategies. The manifold of unitary matrices appears as a collection of all spatial information available in the MIMO wireless systems independent of specific transmission strategies. On these manifolds, signal processing building blocks such as differencing and prediction are derived. Using the proposed signal processing tools on the manifold, this dissertation addresses the CSI coding accuracy, tracking of the CSI under time variation, and compensation techniques for delayed CSI feedback. Applications of the proposed algorithms in single-user and multiuser systems show that most of the spatial benefits of MIMO wireless systems can be harvested.
Mitigation of harmonic and inter-harmonic effects in nonlinear power converters
(2010-12) Cho, Won Jin; Santoso, Surya
Harmonic distortions are inevitably caused by a rectifier and an inverter due to their inherent nonlinearities. An AC-DC-AC converter, configured by the series connection of a rectifier, DC link, and an inverter, induces harmonic distortions at both AC sides and at the DC link. These harmonics can nonlinearly interact or modulate the fundamental frequencies at the AC sides to cause interharmonic distortions. Harmonic and interharmonic distortions can seriously hamper the normal operation of the power system by means of side effects such as excitation of undesirable electrical and/or mechanical resonances, misoperation of control devices, and so forth. This dissertation presents effective methodologies to mitigate harmonic and interharmonic distortions by applying dithered pulse-width modulated (PWM) signals to a voltage-sourced inverter (VSI) type adjustable speed drive (ASD). The proposed methods are also efficient because the dithering applications are performed on control signals without the need for additional devices. By the help of dithering, the rejection bandwidth of a harmonic filter can be relaxed, which enables a lower-order configuration of harmonic filters. First, this dissertation provides a dithering application on gating signals of a sinusoidal PWM (SPWM) inverter in the simulated VSI-ASD model. The dithering is implemented by adding intentional noise into the SPWM process to randomize rising and falling edges of each pulse in a PWM waveform. As a result of the randomized edges, the periodicity of each pulse is varied, which result in mitigated harmonic tones. This mitigation of PWM harmonics also reduces associated interharmonic distortions at the source side of the ASD. The spectral densities at harmonic and interharmonic frequencies are quanti fied by Fourier analysis. It demonstrates approximately up to 10 dB mitigation of harmonic and interharmonic distortions. The nonlinear relationship between the mitigated interharmonics and harmonics is confirmed by cross bicoherence analysis of source- and DC-side current signals. Second, this dissertation proposes a dithered sigma-delta modulation (SDM) technique as an alternative to the PWM method. The dithering method spreads harmonic tones of the SD M bitstream into the noise level. The noise-shaping property of SDM induces lower noise density near the fundamental frequency. The SDM bitstream is then converted into SDM waveform after zero-order interpolation by which the noise-shaping property repeats at every sampling frequency of the bitstream. The advantages of SDM are assessed by comparing harmonic densities and the number of switching events with those of SPWMs. The dithered SD M waveform bounds harmonic and noise densities below approximately -30 dB with respect to the fundamental spectral density without increasing the number of switching events. Third, this dissertation provides additional validity of the proposed method via hardware experiments. For harmonic assessment, a commercial three-phase inverter module is supplied by a DC voltage source. Simulated PWM signals are converted into voltage waveforms to control the inverter. To evaluate interharmonic distortions, the experimental configuration is extended to a VSI-ASD model by connecting a three-phase rectifier to the inverter module via a DC link. The measured voltage and current waveforms are analyzed to demonstrate coincident properties with the simulation results in mitigating harmonics and interharmonics. The experimental results also provide the efficacy of the proposed methods; the dithered SPWM method effectively mitigates the fundamental frequency harmonics and associated interharmonics, and the dithered SDM reduces harmonics with the desired noise-shaping property.
Model-based signal processing for radar imaging of targets with complex motions
(2002) Li, Junfei; Ling, Hao
Model-based signal processing for inverse synthetic aperture radar (ISAR) imaging of targets with complex motions is proposed in this dissertation. Target motion is the most important issue in radar imaging of an unknown target. Although widely recognized as a promising tool in target recognition, ISAR imaging is not yet fully operational in real-world data processing. This is mainly due to the fact that an unknown target, especially a non-cooperative target could have complex motions. First, the performance of existing motion compensation algorithms is evaluated. For this purpose, three sets of radar images of an aircraft, including blind motion compensated images, truth motion compensated images, and predicted images using electromagnetic-code simulation are generated. The limitations of existing radar imaging algorithms are identified after a comparison of the radar images. vii The remaining part of this research focuses on how to overcome these limitations. This is achieved by performing target feature extraction in the presence of complex motions, including three-dimensional (3D) motion, non-rigid body motion and high order motion. For a target with non-planar motion, an algorithm based on the phase analysis of multiple point scatterers is proposed to blindly detect the existence of 3D motion from radar data. An adaptive feature extraction technique is also applied for 3D ISAR image reconstruction from undersampled radar data when the target pose data is known. For a target with non-rigid body motions, adaptive chirplet signal representation is used to first separate signals from the main body and the rotating parts. Better extraction of target geometric features and micro-Doppler features are achieved after individual processing of the separated signal. For a target with high order motions, genetic algorithms are used to replace exhaustive search to reduce the computational time. Throughout the research, the use of physical models is emphasized for better understanding of the radar data. Model-based processing, including adaptive joint time-frequency techniques and genetic algorithms are applied in the information extraction process. Point scatterer simulations are extensively used to test the correctness and to demonstrate the concept of the proposed methods. Results from measurement data are included to demonstrate the effectiveness of the work on realworld problems.
Modular pipeline fast Fourier transform algorithm
(2003-05) El-Khashab, Ayman Moustafa; Swartzlander, Earl E.
A modular pipeline architecture for computing discrete Fourier transforms (DFT) is demonstrated. For an N point DFT, two conventional pipeline √ N point fast Fourier transform (FFT) modules are joined by a specialized center element. The center element contains memories, multipliers and control logic. Compared with a standard N point pipeline FFT, the modular pipeline FFT reduces the number of delay lines required. Further, the coefficient memory is concentrated within the center element, reducing the storage requirements in each of the conventional FFT modules. The centralized memory and address generator provide the data storv age and reordering. The data throughput of a conventional pipeline architecture is maintained with a slightly higher end-to-end latency. The architecture and control logic for both a radix-2 and radix-4 modular pipeline FFT is explained and compared to the traditional pipeline FFT. Further, this methodology facilitates the hardware computation of long FFTs when compared to previous techniques. The new logic developed to control the FFT unit is similar in complexity to current systems and does not rely on any exotic components or hardware features. In fact, the control logic can be reduced to a single counter and a handful of combinational logic. Specifically, using the modular FFT algorithm reduces the overall complexity of the hardware pipeline, permits the use of reusable modules, and does not impact the throughput. The reduction in delay lines lowers the dynamic power consumption. The hardware architecture is particularly suited to reprogrammable and custom devices. Simulations are conducted to analyze the architecture. Experimental results for both radix-2 and radix-4 FFTs are presented and compared with the conventional pipeline FFT. A numerical analysis of the modular pipeline FFT is performed and compared to that of a conventional pipeline FFT.
Multi-objective optimization of antennas for ultra-wideband applications
(2008-05) Kerkhoff, Aaron Jon, 1976-; Ling, Hao
There are a growing number of ultra-wideband applications, which involve the radiation or reception of electromagnetic signals over frequency bandwidths ranging from 1.3:1 to over 10:1. In the design of antennas for ultra-wideband systems, many design objectives must be considered, including impedance matching, radiation efficiency, radiation pattern stability, size, and possibly impulse response. Given the very wide bandwidths considered, it can be challenging to meet all objectives simultaneously, and optimization techniques are useful to achieve a reasonable compromise between objectives. In this dissertation, multi-objective Genetic Algorithm (GA) optimization is used to design ultra-wideband antennas for use in wireless communications and low frequency radio astronomy. GA optimization is first applied to the design of ultra-wide bandwidth planar monopole antennas, which exhibit a narrow-band frequency notch in order to mitigate interference with co-located radio systems. The GA optimizer uses a weighted sum cost function related to impedance matching and radiation patterns at frequencies within both the wide operating band and the narrow notch band to improve antenna performance. A two-dimensional matrix chromosome is used in the GA to represent a wide-range on planar element shapes. It is shown that the GA generates antenna designs which exhibit wideband performance equal to traditional band-notched designs, but have improved azimuth plane radiation pattern symmetry, which widens the effective notch bandwidth. Pareto GA optimization is then applied to the design of planar dipole antenna elements operating over a ground plane for use in a low frequency radio telescope array. The objectives considered include Galactic background or "sky noise reception level, and radiation patterns over the operating band of 20 to 80 MHz. It is demonstrated that the Pareto GA approach generates a set of designs, which exhibit a wide range of trade-offs between the two design objectives, and satisfy all applied geometrical constraints. Multiple GA executions are performed to determine how antenna performance trade-offs are affected by different geometrical constraint values, feed impedance values, radiating element shapes and orientations, and ground conditions. In a follow-up to the previous study, the effects of mutual coupling in a low frequency radio telescope array are considered. It is first shown that a simple receive-based definition of coupling between two antennas can be used to design antenna elements which exhibit reduced mutual coupling effects when operated in a large phased array. This result is utilized in order to perform Pareto GA optimization of wire frame bow-tie dipole elements in terms of mutual coupling, as well as sky noise response and radiation patterns over the 20 to 80 MHz band. The GA generates a set of designs that span a wide range of objective values. The results are analyzed to understand the trade-offs that may be made between the three objectives.
Predicting friction with improved texture characterization
(2017-05) Zuniga Garcia, Natalia; Prozzi, Jorge Alberto
Current methodologies to measure road friction present several disadvantages that make them impractical for field data collection over large highway networks. Thus, it is important to study different ways to estimate surface friction characteristics based on other properties that are easier to measure. The main objective of this study was to analyze surface texture characteristics and to observe their influence on friction. A Line Laser Scanner (LLS) was implemented to make an improved characterization of the road texture which includes macro- and micro-texture description using different texture parameters. Field measurements of friction and texture were collected around Texas using different tests methods. The friction characterization tests included the British Pendulum test (BPT), the Dynamic Friction test (DFT), and the Micro GripTester. Thirty-six different pavement sections were evaluated, including different surface types such as hot-mix asphalt (HMA), surface treatment, and concrete sidewalk. Among the principal conclusions, it was found that there is not a unique relationship between texture and friction. The relationship between texture and friction is strong but it is different for each type of surface, thus, cross-sectional analysis cannot be utilized to quantify the relationship. Additionally, the prediction of friction measures obtained using the BPT and the DFT significantly improved when including information of both macro- and micro-texture into the prediction model. Therefore, a measure of micro-texture should be included into friction models based on texture. Finally, among the study of different texture parameters, the mean profile depth (MPD) was the most significant parameter for macro- and for micro-texture to explain the distinct friction measures.