Signal Processing Concepts Research Articles

Inculcation of Life Philosophy through DSP Concepts

Engineering education is a training that encompass learning the theoretical knowledge of concepts, and real time application domains. The role of engineering is also bound by ethical values and the understanding of social and cultural values. In this context a new approach has been designed to facilitate learning of life philosophies through Digital Signal Processing (DSP) concepts. To facilitate learning process, skits based on core DSP concepts were developed using simple day to day tools and performed with learners. The concept was introduced with simple analogies. Then students were made to participate in skits corresponding to key concepts. After the skits, learners were educated about the correlation with Life philosophies. Three case studies were taken up for this study. We found that the inculcation of life philosophies through simple skits, helped learners better to understand the concepts as well the moral values of life.

PhlatCam: Designed Phase-Mask Based Thin Lensless Camera.

We demonstrate a versatile thin lensless camera with a designed phase-mask placed at sub-2 mm from an imaging CMOS sensor. Using wave optics and phase retrieval methods, we present a general-purpose framework to create phase-masks that achieve desired sharp point-spread-functions (PSFs) for desired camera thicknesses. From a single 2D encoded measurement, we show the reconstruction of high-resolution 2D images, computational refocusing, and 3D imaging. This ability is made possible by our proposed high-performance contour-based PSF. The heuristic contour-based PSF is designed using concepts in signal processing to achieve maximal information transfer to a bit-depth limited sensor. Due to the efficient coding, we can use fast linear methods for high-quality image reconstructions and switch to iterative nonlinear methods for higher fidelity reconstructions and 3D imaging.

Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) network

Because of their effectiveness in broad practical applications, LSTM networks have received a wealth of coverage in scientific journals, technical blogs, and implementation guides. However, in most articles, the inference formulas for the LSTM network and its parent, RNN, are stated axiomatically, while the training formulas are omitted altogether. In addition, the technique of “unrolling” an RNN is routinely presented without justification throughout the literature. The goal of this tutorial is to explain the essential RNN and LSTM fundamentals in a single document. Drawing from concepts in Signal Processing, we formally derive the canonical RNN formulation from differential equations. We then propose and prove a precise statement, which yields the RNN unrolling technique. We also review the difficulties with training the standard RNN and address them by transforming the RNN into the “Vanilla LSTM”11The nickname “Vanilla LSTM” symbolizes this model’s flexibility and generality (Greff et al., 2015). network through a series of logical arguments. We provide all equations pertaining to the LSTM system together with detailed descriptions of its constituent entities. Albeit unconventional, our choice of notation and the method for presenting the LSTM system emphasizes ease of understanding. As part of the analysis, we identify new opportunities to enrich the LSTM system and incorporate these extensions into the Vanilla LSTM network, producing the most general LSTM variant to date. The target reader has already been exposed to RNNs and LSTM networks through numerous available resources and is open to an alternative pedagogical approach. A Machine Learning practitioner seeking guidance for implementing our new augmented LSTM model in software for experimentation and research will find the insights and derivations in this treatise valuable as well.

Graph Moving Object Segmentation

Moving Object Segmentation (MOS) is a fundamental task in computer vision. Due to undesirable variations in the background scene, MOS becomes very challenging for static and moving camera sequences. Several deep learning methods have been proposed for MOS with impressive performance. However, these methods show performance degradation in the presence of unseen videos; and usually, deep learning models require large amounts of data to avoid overfitting. Recently, graph learning has attracted significant attention in many computer vision applications since they provide tools to exploit the geometrical structure of data. In this work, concepts of graph signal processing are introduced for MOS. First, we propose a new algorithm that is composed of segmentation, background initialization, graph construction, unseen sampling, and a semi-supervised learning method inspired by the theory of recovery of graph signals. Second, theoretical developments are introduced, showing one bound for the sample complexity in semi-supervised learning, and two bounds for the condition number of the Sobolev norm. Our algorithm has the advantage of requiring less labeled data than deep learning methods while having competitive results on both static and moving camera videos. Our algorithm is also adapted for Video Object Segmentation (VOS) tasks and is evaluated on six publicly available datasets outperforming several state-of-the-art methods in challenging conditions.

Training a digital model of a deep spiking neural network using backpropagation

Deep spiking neural networks are one of the promising eventbased sensor signal processing concepts. However, the practical application of such networks is difficult with standard deep neural network training packages. In this paper, we propose a vector-matrix description of a spike neural network that allows us to adapt the traditional backpropagation algorithm for signals represented as spike time sequences. We represent spike sequences as binary vectors. This enables us to derive expressions for the forward propagation of spikes and the corresponding spike training algorithm based on the back propagation of the loss function sensitivities. The capabilities of the proposed vector-matrix model are demonstrated on the problem of handwritten digit recognition on the MNIST data set. The classification accuracy on test data for spiking neural network with 3 hidden layers is equal to 98.14%.

Significance of Vocoders in Mobile Communication

Signal Processing finds its applications in many fields of engineering like Communications, Multimedia processing like audio, speech, video compression. It is the constant endeavor in compression techniques to retain highest quality of media through lowest bits so that communication band width utilization can be maximized. The most basic form of end user communication in the mobile industry has been voice calls. Speech signals captured at the microphone are sampled at 8 or 16 KHz and using 16 bits per sample presents 128 KHz of information. Transmitting all of it is very inefficient usage of the communication channel. There are multiple ways to encode the speech input at lower bit rates using Source Coding and Waveform coding techniques. This paper is scoped to practically simulate some of the basic and advanced signal processing concepts and apply them to the speech signal processing domain to minimize the vocoders packet exchange bit rate. Existing techniques are studied and a new schema is proposed which reduce the vocoders packet exchange bit rate further.

Downsampling of Bounded Bandlimited Signals and the Bandlimited Interpolation: Analytic Properties and Computability

Downsampling and the computation of the bandlimited interpolation of discrete-time signals are two important concepts in signal processing. In this paper we analyze the downsampling operation regarding its impact on the existence and computability of the bounded bandlimited interpolation. We assume that the discrete-time signal is obtained by downsampling the samples of a bounded bandlimited signal that vanishes at infinity, and we study two problems. First, we investigate the existence of the bounded bandlimited interpolation for such discrete-time signals from a signal theoretic perspective and show that there exist signals for which the bounded bandlimited interpolation does not exist. Second, we analyze the algorithmic generation of the bounded bandlimited interpolation, using the concept of Turing computability. Turing computability models what is theoretically implementable on a digital computer. Interestingly, it turns out that even if the bounded bandlimited interpolation exists analytically, it is not always computable, which implies that there exists no algorithm on a digital computer that can always compute it. Computability is important in order that the approximation error be controlled. If a signal is not computable, we cannot ascertain whether the computed signal is sufficiently close to the true signal, i.e., we cannot verify every approximation accuracy.

Defense applications of Edmund J. Sullivan’s acoustic signal processing concepts and methods

Passive acoustic signal processing for defense involves the detection, classification, localization and tracking of sound sources. Ed Sullivan posited that “acoustic signals carry information in their kinematics, as well as their spatial and temporal phase structure, so that including information in the signal model will provide enhanced performance.” This idea is applied to the practical defense problem of locating the point-of-origin of hostile sniper fire by processing the acoustic signals sensed by a receiving array. Superior acoustic source localization performance is achieved by including the deceleration of the bullet in the signal model. Next, Ed’s passive synthetic aperture methods are applied to the inverse problem of estimating the source (or rest) frequencies and flight motion parameters of turboprop aircraft and helicopters using a single acoustic sensor and then, an array of sensors. In this case, the source signal is temporally coherent and the source kinematics is contained in the received Doppler-shifted signals. Finally, source localization of broadband signals using cross-correlation methods is observed to fail when the different phase structures of the received signals lead to them being uncorrelated, which can be remedied by differential Doppler compensation.Passive acoustic signal processing for defense involves the detection, classification, localization and tracking of sound sources. Ed Sullivan posited that “acoustic signals carry information in their kinematics, as well as their spatial and temporal phase structure, so that including information in the signal model will provide enhanced performance.” This idea is applied to the practical defense problem of locating the point-of-origin of hostile sniper fire by processing the acoustic signals sensed by a receiving array. Superior acoustic source localization performance is achieved by including the deceleration of the bullet in the signal model. Next, Ed’s passive synthetic aperture methods are applied to the inverse problem of estimating the source (or rest) frequencies and flight motion parameters of turboprop aircraft and helicopters using a single acoustic sensor and then, an array of sensors. In this case, the source signal is temporally coherent and the source kinematics is contained in the received...

Experimental Testing of Vibration Mitigation in 3D-Printed Architected Metastructures

Abstract Band gaps in metamaterials and phononic crystals provide a way to engineer vibration mitigation into a material’s geometry. Here, we present a comprehensive experimental characterization of band gaps in lattice-resonator metastructures, which have been previously analyzed with finite element simulations, to better understand this phenomenon in 3D-printed materials. We fabricate the metastructures with a new approach to obtain multimaterial structures using stereolithography. We experimentally characterize the material’s frequency-dependent storage and loss modulus over the band gap frequencies to confirm that the measured band gaps are due to geometry and not due to material properties. Experimental results using both frequency sweep and impulse excitations show that band gaps and attenuation efficiencies strongly depend on the lattice geometry as well as loading direction, and a comparison between axial and bending excitation responses reveals frequency ranges of “fluid-like” and “optical-like” behaviors. Comparison between finite element simulations and experimental results demonstrate the robustness of the metastructure design. While the experiments used here are well established, their combination allows us to gain additional insights into band gaps measurements. Specifically, we show that the coherence function, a common concept in signal processing, is a strong predictor of band gaps in linear materials and that the attenuation efficiency inside the measured band gap can be physically limited by fluid–structure interactions.

Interpretation of Entropy Algorithms in the Context of Biomedical Signal Analysis and Their Application to EEG Analysis in Epilepsy

Biomedical signals are measurable time series that describe a physiological state of a biological system. Entropy algorithms have been previously used to quantify the complexity of biomedical signals, but there is a need to understand the relationship of entropy to signal processing concepts. In this study, ten synthetic signals that represent widely encountered signal structures in the field of signal processing were created to interpret permutation, modified permutation, sample, quadratic sample and fuzzy entropies. Subsequently, the entropy algorithms were applied to two different databases containing electroencephalogram (EEG) signals from epilepsy studies. Transitions from randomness to periodicity were successfully detected in the synthetic signals, while significant differences in EEG signals were observed based on different regions and states of the brain. In addition, using results from one entropy algorithm as features and the k-nearest neighbours algorithm, maximum classification accuracies in the first EEG database ranged from 63% to 73.5%, while these values increased by approximately 20% when using two different entropies as features. For the second database, maximum classification accuracy reached 62.5% using one entropy algorithm, while using two algorithms as features further increased that by 10%. Embedding entropies (sample, quadratic sample and fuzzy entropies) are found to outperform the rest of the algorithms in terms of sensitivity and show greater potential by considering the fine-tuning possibilities they offer. On the other hand, permutation and modified permutation entropies are more consistent across different input parameter values and considerably faster to calculate.

Connectivity steered graph Fourier transform for motor imagery BCI decoding

Objective. Graph signal processing (GSP) concepts are exploited for brain activity decoding and particularly the detection and recognition of a motor imagery (MI) movement. A novel signal analytic technique that combines graph Fourier transform (GFT) with estimates of cross-frequency coupling (CFC) and discriminative learning is introduced as a means to recover the subject’s intention from the multichannel signal. Approach. Adopting a multi-view perspective, based on the popular concept of co-existing and interacting brain rhythms, a multilayer network model is first built from empirical data and its connectivity graph is used to derive the GFT-basis. A personalized decoding scheme supporting a binary decision, either ‘left versus right’ or ‘rest versus MI’, is crafted from a small set of training trials. Electroencephalographic (EEG) activity from 12 volunteers recorded during two randomly alternating, externally cued, MI tasks (clenching either left or right fist) and a rest condition is used to introduce and validate our methodology. In addition, the introduced methodology was further validated based on dataset IVa of BCI III competition. Main results. Our GFT-domain decoding scheme achieves nearly optimal performance and proves superior to alternative techniques that are very popular in the field. Significance. At a conceptual level, our work suggests a fruitful way to introduce network neuroscience in BCI research. At a more practical level, it is characterized by efficiency. Training is realized using a small number of exemplar trials and decoding requires very simple operations that leaves room for real-time implementation.

THz-to-optical conversion in wireless communications using an ultra-broadband plasmonic modulator

Future wireless communication networks will need to handle data rates of tens or even hundreds of Gbit s−1 per link, requiring carrier frequencies in the unallocated THz spectrum1,2. In this context, seamless integration of THz links into existing fibre-optic infrastructures3 is of great importance to complement the inherent portability and flexibility advantages of wireless networks and the reliable and virtually unlimited capacity of optical transmission systems. On the technological level, this requires novel device and signal processing concepts for direct conversion of data streams between the THz and optical domains. Here, we demonstrate a THz link that is seamlessly integrated into a fibre-optic network using direct THz-to-optical (T/O) conversion at the wireless receiver. We exploit an ultra-broadband silicon-plasmonic modulator having a 3 dB bandwidth in excess of 0.36 THz for T/O conversion of a 50 Gbit s−1 data stream that is transmitted on a 0.2885 THz carrier over a 16-m-long wireless link. Optical-to-THz (O/T) conversion at the wireless transmitter relies on photomixing in a uni-travelling-carrier photodiode. A high-speed wireless THz communication link is seamlessly integrated into a fibre-optic network. The demonstration relies on an ultra-broadband modulator exploiting two-dimensionally localized gap plasmons for direct conversion of the THz signals to the optical domain.

Breaking the boundaries between analogue and digital

breaking the boundaries between analogue and digital

Real-Time Biomagnetic Signal Processing for Uncooled Magnetometers in Cardiology

Despite obvious advantages over their electric counterparts, magnetic measurements are still performed very rarely for medical diagnosis. The high operating costs of established systems with a sufficient limit of detection, mainly based on superconducting quantum interference devices, prevent the spread of magnetic diagnostic procedures. Consequently, the demand for alternative low-cost magnetic sensor systems is large. Several suitable, uncooled sensor concepts exist, but their usability-especially, when operated outside magnetically shielded chambers–requires improvement. This paper highlights basic, real-time signal processing concepts, and implementations to improve the signal quality for different biomedical applications with a focus on cardiology in unshielded measurement environments. Several processing steps in the digital domain are described, such as noise reduction (by means of cancellation and suppression), adaptive sensor signal combination, and adaptive signal averaging. Finally, basic cardiologic feature extraction methods are performed using uncooled magnetometers in combination with the described signal enhancement stages.

SART3D: A MATLAB toolbox for spatial audio and signal processing education

AbstractSignal processing courses form usually a common block within the curriculum of electrical engineering and computer science undergraduate studies. While signal processing is the science behind many technologies used in our everyday life, students usually feel that the practical use of signal processing subjects is obscured due to their intensive theoretical perspective. In this context, audio signal processing is a discipline where many general signal processing concepts can be easily experienced with sound. Specifically, spatial audio, i.e., the manipulation of audio signals with the aim of capturing and/or recreating the spatial impression of sound, provides an excellent and appealing application framework for signal processing education. This paper presents SART3D, an open‐source MATLAB toolbox that facilitates spatial audio and signal processing education by enabling students to focus on the implementation and testing of core algorithms and to experience their own results using real‐time processing. The time required for other nonrelevant aspects, such as hardware and driver set‐up or user interaction, is significantly reduced. Experience evaluated throughout four academic years and 45 students (25% female) indicate they felt particularly engaged when SART3D was included in the course, improving their performance and understanding of the subject. In this context, the four factors defined by the Student Course Engagement Questionnaire (skills, emotional, participation, and performance) were analyzed to assess engagement in its multiple dimensions.

Mathematical Model of Multirate Signal Processing Concept on down/up Sampling Multistate and Polyphase Decompositions for DSP and Digital Communication Application

Mathematical Model of Multirate Signal Processing Concept on down/up Sampling Multistate and Polyphase Decompositions for DSP and Digital Communication Application

Revisiting signal processing with spectrogram analysis on EEG, ECG and speech signals

Abstract Biomedical signal processing is the utilization of digital signal processing techniques, such as Fourier transform, filtering, spectral estimation and wavelet transform to biomedical complications, such as the analysis of breathing cycle, cardiac signals, brain signals, etc. Digital filters are used to preserve the in-band signals and to block out-of-band noise. Low-pass, high-pass, band-pass, and band-stop filters are commonly used for filtering applications. The digital signal processing (DSP) concepts can be used for other biomedical applications such as biomedical imaging (MRI, CT, X-ray, PET, ultrasound) and genomic signal processing too. The experiments such as denoising corrupted ECG signals, respiratory artifacts removal, and an identification of rhythmic patterns from EEG signals with spectrogram analysis are performed. Matlab R2016b tool is used for this study. Finally, the assessment, survey and feedback results from the student are tabulated for the better improvement of the course study.

Development of an in-air circular synthetic aperture sonar system as an educational tool

Synthetic aperture sonar (SAS) is an underwater acoustic imaging technique that is used to generate high quality imagery through the coherent combination of signals measured on a moving platform. The practical challenges of underwater experimentation and the lack of publicly available data sets make it difficult to teach SAS processing to students using experimental data. These challenges are compounded by the complicated signal processing that is required to compensate for the unwanted platform motion that is common in field data. This paper details the development of an in-air circular SAS system using commonly available laboratory equipment. By operating in air and using a circular geometry many of the challenges that students would otherwise encounter in SAS processing are either well controlled or completely eliminated. Imagery collected on the system is shown to demonstrate how meaningful data can be generated using a simple set of equipment. A number of applications of the system as a tool to teach array signal processing, pulse compression, and imaging concepts are also discussed.

Introduction to the Special Issue on Array Signal Processing for Angular Models in Massive MIMO Communications

The papers in this special issue focus on array signal processing for angular models in massive MIMO communications. In recent years, there has been tremendous research progress in massive MIMO from communications theory viewpoint. An emerging trend is to explore the angular models of the propagation channels and design the massive system with array signal processing techniques. Particularly, fundamental concepts in array signal processing (e.g., direction of arrival/departure - DOA/DOD), can be conceptualized and exploited.

Signal Estimation & Reproduction in Frequency Domain using Tunable Q-Factor Wavelet Transform

The present patterns and development in communication and digitization builds the importance of signal and image processing concepts. This paper concentrates on former one i.e., signal processing, and mainly concentrates on receiver section in planning and actualizing the channels to recuperate the information with less measure of noise or with high SNR. In this paper, rather than processing the time domain data, spectral data nothing but frequency domain processing is executed on the approximate coefficients produced with the backing of Tunable Q-Factor Wavelet Transform [TQWT].