Search

This paper deals with the issue of finite-time observer-based H ∞ control design of networked switched systems (NSS) subject to external disturbances via an event-triggered observation strategy. This triggering strategy is used to decide when the output measurements are sent to the observer over the communication network. By incorporating digital networks, the closed-loop system is modeled as a switched delayed system due to network-induced delays. By using the average dwell time (ADT) switching method with the Lyapunov–Krasovskii functional, several sufficient linear matrix inequalities (LMIs) conditions for the observer-based control synthesis are developed to guarantee the finite-time boundedness and the disturbance attenuation control performance of the resulting switched delayed system. Finally, an example is given to illustrate the effectiveness of the theoretical results. [ABSTRACT FROM AUTHOR]

Since the coprime discrete time fractional Fourier transform (DTFrFT) spectral receiver can effectively intercept low probability of interception signals and its output shows favorable spectral characteristics in the discrete time fractional Fourier domain (DTFrFD), this paper presents an intrapulse modulation recognition method in the DTFrFD to process the coprime DTFrFT spectral receiver's output data, consisting of two banks, namely coarse recognition and fine recognition. The coarse recognition returns a coarse classification for the shift keying signals [binary phase shift keying (BPSK) and binary frequency shift keying (BFSK)] from the linear frequency modulation (LFM) and polyphase coded signals (Frank code, P1 code, P2 code, P3 code, P4 code) by addressing the optimal transformation order of the intercepted data in DTFrFD. In the fine recognition algorithm, the ridge line extraction function and the peak energy ratio function in the DTFrFD are defined to identify BPSK signals from BFSK signals and distinguish LFM signals and polyphase-coded signals. Compared to conventional recognition approaches in the Fourier domain, the proposed recognition method is always more efficient because it utilizes more spectral characteristics of the output data in the DTFrFD. Both theoretical analysis and simulation results demonstrate that the proposed algorithm shows better robustness and better recognition performance than the original coprime digital Fourier transform spectral receiver against SNR variation. [ABSTRACT FROM AUTHOR]

In this paper, a ternary true random number generator (TTRNG) is designed and simulated using the stochastic behavior of the magnetic tunnel junction (MTJ) device at currents lower than the critical current and the adjustability of the threshold voltage carbon nanotube field effect transistors. The stochastic behavior of the MTJ ensures the generation of a truly random sequence, while the CNTFET threshold voltage adjustability enables the design and implementation of the ternary circuits. To ensure the equality of the ratio of numbers generated by the designed TTRNG and to avoid the influence of fabrication process variation, a post-processing block has been designed and used in the proposed TTRNG. The results of the simulations show that the proposed TTRNG occupies at least 47% lower area and consumes 42% lower power than the state-of-the-art counterparts. Also, the statistical analysis results show that the variation in the ratio of random numbers generated even in fabrication process variation is less than 5%. [ABSTRACT FROM AUTHOR]

In this paper, a new robust adaptive beamforming method based on the autoregressive (AR) power spectrum is proposed. To improve the robustness of the Capon spectrum, the AR model is applied to realize the desired signal power and interference power estimations, which are used to reconstruct the covariance matrix. Besides, the eigenvalue decomposition is used to remove the redundancy of the reconstructed interference-plus-noise covariance matrix, where the number of interferences is confirmed by the maximum ratio index of the adjacent eigenvalues. Numerical simulations highlight that the proposed method is more robust against some common errors compared with several beamformers. [ABSTRACT FROM AUTHOR]

In this paper, the finite-time control of nonlinear impulsive switched positive systems (ISPSs) is studied, where the impulses and bounded disturbance are both fully considered. By designing a novel event-triggered strategy, we present LMI-based conditions for the existence of a controller that guarantees finite-time boundedness (FTB) of the resulting closed-loop systems based on the average dwell time (ADT) method. Further, the finite-time weighted L 2 -gain performance of the considered systems is analyzed. Also, a lower bound of the minimum inter-event interval is obtained to preclude the Zeno behavior. Finally, a numerical example is given to compare the differences between different controllers and to verify the effectiveness of the presented conclusions. [ABSTRACT FROM AUTHOR]

In this paper, we study the speech emotion feature optimization using stochastic optimization algorithms, and feature compensation using deep neural networks. We also proposed to use accentuation-based fusion for long-time speech emotion recognition. Firstly, the extraction method of emotional features is studied, and a series of speech features are constructed for the recognition of emotion. Secondly, we propose a method of sample adaptation through denoising autoencoder to enhance the versatility of features through the mapping of sample features to improve adaptive ability. Thirdly, GA and SFLA are used to optimize the combination of features to improve the emotion recognition results at the utterance level. Finally, we use transformer model to implement accentuation-based emotion fusion in long-time speech. The continuous long-time speech corpus, as well as the public available EMO-DB, are used for experiments. Results show that the proposed method can effectively improve the performance of long-time speech emotion recognition. [ABSTRACT FROM AUTHOR]

In recent times, the use of chest X-ray (CXR) images in telemedicine has increased exponentially. However, the transfer of these images on an unsecured channel poses a serious threat to their authenticity and copyright protection. To address this problem, a chest X-ray image watermarking scheme (CXRmark) is presented in this paper using an online sequential reduced kernel extreme learning machine (OS-RKELM). U-Net is used to segment the lung area into the region of non-interest (RONI) and region of interest (ROI). An OS-RKELM is trained using the approximation coefficients (obtained by transforming training images using discrete wavelet transform) and the corresponding quantized coefficients. Subsequently, it is used to modulate the approximation coefficients. In order to avoid inadvertent deterioration in the ROI and to achieve high robustness, the proposed watermarking scheme uses different embedding strengths for ROI and RONI. Further, to enhance the security of the watermark, it has been encoded using the logistic chaotic map. To evaluate the performance of the proposed scheme, we experimented with a set of 461 CXR images from a publicly available repository: COVID-19 Image Data Collection. It has been demonstrated that CXRmark results in high perceptual quality of the signed CXRs and is robust to common interference. Further, the reversibility of the proposed CXRmark scheme is demonstrated by recovering the undistorted CXR image from the signed CXR. The comparison of the proposed scheme with state-of-the-art schemes shows that CXRmark outperforms its competitors in terms of imperceptibility and robustness. [ABSTRACT FROM AUTHOR]

The paper discusses in detail the inability of a circuit designer to design a single input single element-controlled universal memelement emulator. A Single input Single element-controlled universal memelement emulator would be such a configuration that can provide the realization of any of the three memelements (Memristor/Memcapacitor/Meminductor) by taking a circuit element as inductor (L), Capacitor (C) or Resistor (R) with the same input port. The analysis shows that no such circuit configuration can be built to realize such a highly flexible universal memelement emulator. It is found that by using such structures, only two ideal memelements can be realized at maximum. If these circuit structures are used to realize the remaining third memelement, then, the resulting element comes out to be a non-ideal memelement. Two circuit configurations to demonstrate this theory are also included for both charge-controlled and flux-controlled memelement emulation based on the approaches shown in this paper. [ABSTRACT FROM AUTHOR]

This paper presents a high imperceptible and robust audio watermarking algorithm by optimizing the scaling parameter as per the imperceptible requirements with minimum bit error rate. The spread spectrum-based audio watermarking is used in this paper, where the scaling parameter is optimized to achieve trade-off between imperceptibility and robustness of watermarked audio. Primarily, the effect of scaling parameter on the perceptual quality of watermarked audio is investigated and the total error introduced in the host audio due to embedding the watermark is computed. The scaling parameter is optimized for maximizing the robustness by considering objective difference grade (ODG) score (for music signals), perceptual evaluation of speech quality (PESQ) score (for speech signals) as a constraint to meet the imperceptibility requirements. To solve this proposed optimization problem, two search algorithms are developed. The embedding is performed in low-pass framelet transform coefficients through SVD with the optimized scaling parameter. The experimental results show that the proposed algorithm achieves good imperceptibility with an average ODG score of −0.32 and PESQ score of 3.86 for music and speech signals, respectively, under various payload conditions. The proposed algorithm shows better robustness to the common signal processing attacks such as noise addition, filtering, resampling, MP3 compression, amplitude scaling, cropping, and requantization. [ABSTRACT FROM AUTHOR]

This paper presents a novel power and energy-efficient carbon nanotube field-effect transistor (CNTFET)-based design of a quaternary magnitude comparator for single and double digits. It also proposes an improved circuit of the ternary single-digit comparator and a novel design of the double-digit ternary comparator. The designs make use of the pass transistor logic and transmission gates, resulting in a more power and energy-efficient design with a lower number of CNTFETs when compared with existing designs. Besides, this paper also proposes a generalized circuit of magnitude comparator for multiple digits. The proposed circuits are simulated in HSPICE using the 32 nm CNTFET model presented by Stanford University. Average power and propagation delays are observed and the power-delay-product (PDP) is calculated in each case. The impact of process parameter variations has been analyzed through Monte Carlo simulation and has been presented in the form of two-dimensional graphs. [ABSTRACT FROM AUTHOR]

Image processing algorithms are essential for clarifying the image and improving the ability to recognize distinct characteristics of the image. The field of digital image processing is widespread in several research and technology applications. In many of these applications, the existence of impulsive noise in the obtained images is one of the most frequent problems. The median filter is a strong method to remove the impulsive noise; it effectively eliminates salt and pepper noise from the image. The main target of this paper is to investigate efficient median filter units to be connected to a general-purpose processor (GPP) for FPGA-based embedded systems. The paper exposes three novel techniques, two of them specially for median filtering techniques and the third one is used to get the maximum number of any 9 elements array. The proposed algorithms are inspired by the Median Of Median (MOM) algorithm. The first two techniques are tested for filtering 3 × 3 image windows and optimized for producing the expected result in high accuracy, short time, and reduced number of comparisons. The last technique is tested for a 9 elements array for extracting the maximum number in same high efficiency manner. Furthermore, the three proposed techniques are implemented leveraging the advantage of the parallel processing and the FPGA flexible resources to satisfy the real-time processing constraints. A comparison between the first two proposed filtering units and their counterparts in the literature is included. The comparison reveals the superiority of the first technique in terms of accuracy with fewer comparators than previously published techniques. Besides, the paper illustrates how the concept beyond the proposed techniques can be used to perform the maximum pooling for convolution neural networks. [ABSTRACT FROM AUTHOR]

Multivariable Hammerstein time-delay (MHTD) systems have been widely used in a variety of complex industrial systems; thus, it is of great significance to identify the parameters of such systems. The MHTD system is difficult to identify due to its inherent complexity. As one of heuristic algorithms, the gravitational search algorithm is suitable for identifying such complex models, but it has the problem of easily falling into local optimum. Therefore, this paper combines the improved chaotic gravitational search algorithm (ICGSA) and gradient iterative (GI) algorithm and proposes an ICGSA–GI algorithm to overcome the shortcomings of the above two algorithms. Then, we use it to identify the unknown parameters of the MHTD systems. Finally, a numerical example and an application case are given for validating the feasibility of the three identification methods. The results demonstrate that the three algorithms can identify the unknown parameters of the MHTD system effectively. In addition, by comparing with ICGSA and GI algorithms, this paper confirms that the ICGSA–GI algorithm behaves better than ICGSA and GI algorithm in identification accuracy, and the convergence speed of the ICGSA–GI algorithm is faster than that of the GI algorithm. [ABSTRACT FROM AUTHOR]

Environment Sound Classification (ESC) has been a challenging task in the audio field due to the different types of ambient sounds involved. In this paper, we propose a method for the ESC tasks based on the CAR-Transformer neural network model, which includes stages of sound sample pre-processing, deep learning-based feature extraction, and classifier classification. We convert the one-dimensional audio signal into two-dimensional Mel Frequency Cepstral Coefficients (MFCC) and use them as the feature map of the audio. The CAR-Transformer model was used for feature extraction, and after dimensionality reduction of the extracted feature map, we use the fully connected layer as a classifier of the feature map to obtain the final results. The method achieves a classification accuracy of 96.91% on the UrbanSound8K dataset, while the number of parameters in the model is only 0.16 M. The results of this paper were compared with other state-of-art research. [ABSTRACT FROM AUTHOR]

This paper presents a highly configurable and low-complexity CNN accelerator based on the MobileNetV3 model. To the best of authors' knowledge, this is the first design of CNN accelerator based on the MobileNetV3 model. A highly efficient processing flow and memory-access scheme are proposed in this paper so that the throughput is greatly enhanced for the structural features in MobileNetV3 model. Furthermore, the proposed processing flow enhances the efficiency for the utilization of hardware components to reduce the complexity. Based on the proposed processing flow, this paper presents a highly configurable architecture to support various operation modes in MobileNetV3 model. The designed architecture is synthesized and layout with TSMC 90 nm technology. The evaluations for the performance and area complexity are conducted based on the post-layout estimations. It is shown in this paper that the performance of 197.7 FPS is achieved with the hardware complexity of 5392 KGEs for the MobileNetV3-Large. Compared to the state-of-the-art accelerator based on MobileNet, the FPS of the proposed design is improved by 3.4 × and the complexity is reduced by 18%. [ABSTRACT FROM AUTHOR]

The paper proposes a smart method to remove network nodes that are highly prone to impulse noise and replacing them with a linear combination of existing more reliable nodes. This is done to reduce communication cost. To reduce the communication cost in a network, a common way is to remove some nodes randomly and then replace the intermediate estimation of these nodes by the corresponding node estimation. In this direction, the contribution of this paper is twofold. First, we suggest to remove the nodes smartly by omitting the unreliable nodes prone to impulsive noise. This is done by computing the disturbance induced in the adaptation step of the diffusion least mean square. Second, we replace the estimation of removing nodes by a linear combination of existing estimations of reliable nodes instead of just replacing by the estimation of the corresponding node. Also, the coefficients of linear combination are optimized based on the minimum disturbance principle. Furthermore, the minimum achievable disturbance is calculated mathematically and a necessary and sufficient condition for stability of the proposed algorithm is presented. Finally, the simulation results show the efficiency of the proposed method in comparison with some state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

To solve the problem that the least mean square (LMS) algorithm cannot balance the convergence rate and steady-state mean square error (MSE) well and require excessive manual adjustment of parameters, this paper proposes a parameter-free variable step-size LMS algorithm based on the hyperbolic tangent function. First, the algorithm uses the hyperbolic tangent function to establish a nonlinear relationship between the error and the step size. Based on this, the average value of the error signal is used to update the step size. Further, the accumulated error value is multiplied by its average value and added to the input signal energy. Then, the algorithm is normalized by using this value to prevent the algorithm divergence due to the sudden increase in the signal power. The convergence proof of the proposed algorithm is provided in this paper. The simulation results indicate that the proposed algorithm can automatically adjust the parameters. The convergence rate and the steady-state MSE are better than those of other algorithms for both high and low signal-to-noise ratios. [ABSTRACT FROM AUTHOR]

In this paper, a high-resolution time–frequency (TF) analysis method, called time-extracting wavelet transform (TEWT), is introduced to analyze impulsive-like signals whose TF ridge curves are nearly perpendicular to the time axis. For impulsive-like signals, the instantaneous frequency with almost infinite rate of change is difficult to estimate, but the group delay (GD) with nearly zero rate of change is easier to estimate. Since the GD is the key feature of frequency-domain signals, it indicates that one can try to understand impulsive-like signals from the perspective of frequency-domain signals. In this regard, for an impulsive signal and its Fourier transform (i.e., the frequency-domain harmonic signal), we propose the TEWT that achieves highly concentrated TF representations while allowing signal reconstruction, only by retaining the TF coefficients closely related to TF features of signals, while removing weakly related TF information. The two contributions of this paper are the proposal of TEWT and the theoretical analysis of TEWT for frequency-domain signals. On the other hand, we provide a rigorous theoretical analysis of TEWT under a mathematical framework for frequency-domain signals. Specifically, we define a function class as a set of all superposition of well-separated frequency-domain harmonic-like functions, where each function can be locally regarded as a sum of a finite number of harmonic signals in the frequency domain, and establish error bounds for WT approximate expression, GD estimation, and component recovery. Finally, we verify the effectiveness of TEWT in terms of the energy concentration, robustness, and invertibility through numerical experiments with synthetic and real signals. [ABSTRACT FROM AUTHOR]

The necessary and sufficient conditions for decomposing a fourth-order (FO) linear time-variant system (LTVS) in the form of cascade connection of a first-order system and a third-order system that are commutative are expressed explicitly in this paper. Some additional conditions are also derived for a valid decomposition when the initial conditions are different from zero. Further, explicit formulas are derived for the determination of the decomposed subsystems. The paper also investigated the effects of changing the connection order on the sensitivity and disturbance properties. Theoretical results about the decomposition are supported and illustrated by considering various examples. [ABSTRACT FROM AUTHOR]

In this study, necessary and sufficient conditions for the decomposition of any second-order discrete-time linear time-varying system as a commutative pair of two first-order systems are presented. Commutativity conditions with zero initial conditions are first expressed by Theorem 1. Then, the conditions under nonzero initial conditions are studied and presented by Theorem 2. The results including decomposition formulas are well verified by examples worked by using MATLAB Simulink tool. The importance of the paper subject is emphasized in view of some engineering applications for getting better performance characteristics such as sensitivity, robustness and stability. In fact, the results of this paper can be used when the synthesis of a second-order system is made by cascaded and commutative pairs of simple first-order subsystems; and by using the explicit decomposition formulas derived in the paper this can be achieved very easily. Since different decompositions lead to different performance characteristics, one should choose the appropriate ones among them. [ABSTRACT FROM AUTHOR]

In this paper, the design of event-triggered sliding mode controller for singular Markovian jump systems is studied. Firstly, the event-triggered sliding mode control (ETSMC) law is designed, it can ensure the reachability of sliding surface. Then for singular Markovian jump systems, a lemma is given to overcome the difficulties arising from the ETSMC strategy. Secondly, a set of sufficient condition is proposed by using Lyapunov function method to ensure the stochastic admissibility of the sliding mode dynamics. Then, a positive lower bound of the inter execution time can be ensured, which illustrates that the Zeno behavior does not occur. Finally, the simulation is given to reflect the effectiveness of theoretical results in this paper. [ABSTRACT FROM AUTHOR]

In this article, we introduce the concept of the windowed octonion Fourier transform (WOFT) by taking the octonion-valued function as the window function on the space of square integrable octonion-valued functions on R 3 . Some properties of the windowed octonion Fourier transform (WOFT) like left linearity, parity, specific shift, inversion, orthogonality and Hausdorff–Young inequality were also established. Towards the culmination of this paper, we establish the Pitt's inequality and hence some uncertainty principle for the proposed transform. Some potential applications were also added to show the effectiveness of this paper. [ABSTRACT FROM AUTHOR]

Full adders (FAs) are the core elements and substantially impact the performance of digital signal processing applications such as arithmetic logic unit (ALU). In this paper, XOR gate-based FA design is presented that can operate appropriately in the domain of ultra-low voltage (LV) and low power (LP). In this treatise, FGMOS technique is used to elevate the performance in terms of design complexity and to reduce the power requirements. Important device performance characteristics such as power (pwr), delay (tp), power delay product (PDP) and energy delay product (EDP) are explored for the proposed FGMOS XOR gate design, and existing XOR designs are used to assess the results. This research paper presents a low-power FGMOS XOR gate design with a total power consumption of 5.39 pW at 0.7 V supply. Further, in this work, 1-bit FGMOS-based ALU is proposed to perform the functions of full adder, NAND, NOR and XOR. To ameliorate the performance of the circuit, FGMOS-based NAND and NOR gate designs are also proposed in this treatise. Aforementioned full adder and XOR gate along with pass-transistor logic (PTL)-based multiplexers are used to design efficient design of 1-bit ALU at 22-nm technology node. This research paper explicates the utility of FGMOS technique for the designing of high-performance complex digital circuits. [ABSTRACT FROM AUTHOR]

In this paper, we propose a general purpose, simple and fast fusion algorithm based on guided image filter. The proposed method can well combine useful source image information into the fused image supported by multi-scale image decomposition, structure transferring property, visual saliency detection, and weight map construction. Multi-scale image decomposition is appropriate to represent and manipulate image features at various scales. Structure transferring property enabled by our algorithm can induce structures of one source image into the other. A new visual saliency detection based on guided image filter introduced in this paper is able to extract significant regions from visually different images of the same scene. The choice of weight maps helped to integrate the complementary information pixel by pixel at each scale. Experimental outcomes of the proposed method are compared and analyzed with traditional and recent guided image filter-based fusion algorithms in terms of visual quality, fusion metrics and run time. In addition, to enhance fusion results further we made an effort to find a suitable image and video enhancement algorithm. The fusion performance analysis clearly indicates that the proposed method is very promising along with less run time. [ABSTRACT FROM AUTHOR]

As a sub-Nyquist sampling structure based on compressed sensing (CS), modulated wideband converter (MWC) was proposed to overcome the sampling rate limitation of analog-to-digital converters (ADCs) for wide spectral multiband signals. Different from ideal conditions when analyzing its theory, in practice, the MWC system is implemented with non-ideal devices. Due to this deviation between the ideal model and practical hardware, any imperfection of the analog device will lead to an imperfect signal reconstruction. Although the non-ideal factors can be compensated and calibrated, the conventional compensation and calibration methods easily introduce additional phase errors. This paper aims to evaluate the impact of those phase errors on signal reconstruction performance. We point out that phase errors deteriorate the normalized mean squared errors (NMSE) and even lead to incorrect support set recovery in serious cases. We thus propose a method using the SINC signal and DC signal to measure the MWC channel response which avoids the phase errors introduced in the compensation process. To verify this method, an MWC system and a measuring platform are implemented on hardware. The measurement results show that the recovery NMSE of − 13.09dB is achieved after being measured and compensated with an input SNR of 25.47dB. [ABSTRACT FROM AUTHOR]

The standard proportionate normalized least mean square (PNLMS) adaptive algorithm suffers from convergence performance limitation due to a constant step-size during the convergence period. In this paper, a switching-based variable step-size PNLMS is proposed to improve the convergence performance in sparse system identification. To adjust the step-size, the convergence performance of PNLMS is first analysed in the statistical sense and by exploiting the analysis, a switching-based method is then proposed, which brings about a fast convergence towards the desired steady-state mean-square weight deviation. The step-size reduces during the convergence period in a few steps, while in the case of abrupt change in the system impulse response, the step-size increases to its initial value. A sub-band version of the proposed adaptive algorithm is further proposed for highly correlated input signals. Simulation results confirm the superiority of the proposed full-band and sub-band algorithms in terms of convergence performance compared to some competing adaptive algorithms. [ABSTRACT FROM AUTHOR]

As a combination of sound event detection and direction of arrival, the joint task of sound event localization and detection (SELD) is an emerging audio signal processing task and is applied in many areas widely. A popular convolutional recurrent neural network (CRNN)-based method uses convolution neural network (CNN) to extract high-level space features from manually designed features and utilizes recurrent neural network to model sequence context information. Some studies have shown that the normal CNN could not be robust in challenging acoustic environments such as overlapping, moving and discontinuous sources. To improve the performance of SELD in more complex acoustic scenes, parallel multi-attention enhancement (PMAE) is proposed as a convolution enhancement method to boost the representation ability of CNN in this paper. PMAE consists of attention feature enhancement (AFE) and parallel multi-attention (PMA) block. PMA, embedded into AFE, extracts boosting global–local features by efficient attention modules along with different dimensions. AFE, as a feature fusion strategy, fuses multi-scale enhanced features to improve feature representation. AFE shows great performance for overlapping sources. PMA adequately extracts characteristic information of different sound events and shows better performance on moving and discontinuous sources when it is combined with AFE. Based on such a framework, the SELD system becomes robust, while the target sources are moving and overlapping with unknown interference classes. The simulations show that proposed PMAE improves the performance enormously for SELD without other technologies, such as data augment and post-processing. [ABSTRACT FROM AUTHOR]

The fractional Fourier transform (FrFT) is a useful mathematical tool for signal and image processing. In some applications, the eigendecomposition-based discrete FrFT (DFrFT) is suitable due to its properties of orthogonality, additivity, reversibility and approximation of continuous FrFT. Although recent studies have introduced reduced arithmetic complexity algorithms for DFrFT computation, which are attractive for real-time and low-power consumption practical scenarios, reliable hardware architectures in this context are gaps in the literature. In this paper, we present two hardware architectures based on the referred algorithms to obtain N-point DFrFT ( N = 4 L , L is a positive integer). We validate and compare the performance of such architectures by employing field-programmable gate array implementations, co-designed with an embedded hard processor unit. In particular, we carry out computer experiments where synthesis, error and latency analyses are performed, and consider an application related to compact signal representation. [ABSTRACT FROM AUTHOR]

This paper considers the output-feedback-based proportional–integral–derivative (PID) control and filtering problems for the discrete-discrete Roesser system with disturbances. A PID output-feedback controller, where the integral loop is with fixed time-window, is first constructed for the two-dimensional system. By means of co-positive Lyapunov function approach, sufficient conditions are proposed to ensure that the resulting closed-loop system is positive, exponentially stable and has an l 1 -gain bound γ . In addition, the gains of desired PID controller are then appropriately parameterized by solutions to certain linear programming problems. An illustrated example is provided to show effectiveness of the PID controller designed. [ABSTRACT FROM AUTHOR]

Identifying a scene based on the environment in which the related audio is recorded is known as acoustic scene classification (ASC). In this paper, a bi-level light-weight Convolutional Neural Network (CNN)-based model is presented to perform ASC. The proposed approach performs classification in two levels. The scenes are classified into three broad categories in the first level as indoor, outdoor, and transportation scenes. The three classes are further categorized into individual scenes in the second level. The proposed approach is implemented using three features: log Mel band energies, harmonic spectrograms and percussive spectrograms. To perform the classification, three CNN classifiers, namely, MobileNetV2, Squeeze-and-Excitation Net (SENet), and a combination of these two architectures, known as SE-MobileNet are used. The proposed combined model encashes the advantages of both MobileNetV2 and SENet architectures. Extensive experiments are conducted on DCASE 2020 (IEEE AASP Challenge on Detection and Classification of Acoustic Scenes and Events) Task 1B development and DCASE 2016 ASC datasets. The proposed SE-MobileNet model resulted in a classification accuracy of 96.9% and 86.6% for the first and second levels, respectively, on DCASE 2020 dataset, and 97.6% and 88.4%, respectively, on DCASE 2016 dataset. The proposed model is reported to be better in terms of both complexity and accuracy as compared to the state-of-the-art low-complexity ASC systems. [ABSTRACT FROM AUTHOR]

Imperceptibility, embedding capacity and robustness are basic requirements of an image watermarking algorithm. There is always a trade-off among these requirements. To achieve a balance between these three properties in an image watermarking scheme, this paper presents a high capacity and encrypted dual image watermarking in YCbCr color space. This scheme is based on redundant discrete wavelet transform (RDWT), randomized singular value decomposition (RSVD), Manta ray foraging optimization (MRFO) and Henon mapping. Further, MRFO is utilized to get the optimized strength factor for embedding the watermark into the cover image to achieve high robustness and imperceptibility along with desired capacity. The luminance component of cover image is used to embed the watermark as it contains the vital information. Performance of the proposed scheme is evaluated for various medical and non-medical images in terms of imperceptibility and robustness. Different quality metrics are utilized to assess the performance of scheme. Various attacks are applied to check the robustness of the proposed method. Imperceptibility and robustness along with embedding capacity and computational complexity comparison with existing techniques are also presented. The comparison of MRFO with particle swarm optimization (PSO) is also carried out. Peak signal-to-noise ratio (PSNR) for all cover images is above 60 dB whereas Structural similarity index measure (SSIM) and Normalized correlation (NC) values are above 0.99 under no attacks. For non-medical images, the average percentage improvement in PSNR is 13.76% and in NC is 4.3% in comparison with other existing schemes. In case of medical images, average percentage improvement in PSNR is 41.35% and in NC is 0.62% when compared with related works. The proposed scheme is less computational complex in comparison with other existing schemes. [ABSTRACT FROM AUTHOR]

A multi-dimensional signal on graph has different features along different directions/dimensions. For example, image signal is isotropic along horizontal and vertical directions, and time-varying signal presents different correlation characters in the time and vertex domains. Nevertheless, the current graph signal processing concentrates on the single shift, which makes it difficult to differentiate the correlation features along different directions. In this paper, we define a multi-shift notion for the time-varying signals on graphs enabling us to separately analyze the multi-dimensional signal along different directions captured by diverse shifts. Furthermore, the multi-shift notion is leveraged to formulate the inpainting problem for time-varying signals on the strong product graph, which can be exploited to characterize three different kinds of elements interaction of time-varying signals by using three different kinds of shifts. The conjugate gradient descent algorithm is further deployed to solve the inpainting problem. Numerical experiments conducted on the synthetic signal and real-world data show the potentiality of the multi-shift representation and the effectiveness of the inpainting algorithm. [ABSTRACT FROM AUTHOR]

This paper presents a low-power multi-rate clock and data recovery (CDR) for receivers of serial links. Its basic structure includes a current-mode logic bang–bang phase detector sampled by low-mismatch half-rate quadrature clocks, which are generated by voltage-controlled delay line (VCDL) and two-stage time-average circuits. The total delay of VCDL can be adjusted to accommodate a wide frequency range by its bias voltage, which is generated by a delay-locked-loop-based bias generator. The quadrature clocks are 64-phase adjustable with high linearity, which is realized by phase interpolator with a compensating structure. The parameters of phase detection loop are well designed to satisfy both high jitter tolerance and low clock jitter. Fabricated in a 40 nm CMOS technology, the CDR occupies an active area of 0.036 mm 2 only. With a wide operating range of 2–20 Gb/s, the chip consumes 62.5 mW, corresponding to an energy efficiency of 3.1 pJ/bit. The measured root-mean-square jitter and peak-to-peak jitter for the recovered clock at 9 GHz are 1.9 and 10.8 ps, respectively. [ABSTRACT FROM AUTHOR]

The digital media shared over the Internet is the backbone of the current generation of communication. Multimedia is becoming increasingly popular with no governing body for intellectual property rights, hence the need for protecting it efficiently while preserving the feature set of given data. Several watermarking methods have been proposed for the copyright protection of audio and speech signals yet the Recording Industry Association of America reported illegal downloading costs around $12.5 billion in the year 2021. Furthermore, the challenge to find the right balance between the three contrasting characteristics of watermarking; robustness, imperceptibility, and payload persists. This paper puts forward a new, hybrid, and purely data-driven way of watermarking the audio and speech signals using iterative filtering and singular value decomposition. The input audio is processed using iterative filtering to generate a band of intrinsic mode functions (IMFs) and a residue component. Depending upon the application, singular value decomposition is applied on residue or an IMF for insertion of bits of the watermark to minimize the effect of watermarking on the audio signal. The method focuses on the robustness of the audio signals against various signal processing attacks while providing a significant boost in imperceptibility and payload as well. Thus, providing a strikingly excellent balance between the characteristics of watermarking. With an average SNR of 38.06 db and normalized correlation close to unity, the proposed method validates its superiority compared to the state of the art. [ABSTRACT FROM AUTHOR]

This paper focuses on iterative parameter estimation methods for a nonlinear closed-loop system (i.e., a nonlinear feedback system) with an equation-error model for the open-loop part. By applying negative gradient search, a gradient-based iterative algorithm is constructed. To reduce the computational costs and improve the parameter estimation accuracy, the hierarchical identification principle is employed to derive a hierarchical gradient-based iterative algorithm. A simulation example is provided to test the effectiveness of the proposed algorithms. [ABSTRACT FROM AUTHOR]

In this paper, a blind semi-fragile dual watermarking system for video content authentication and tampering localization is proposed. In this method, two watermarks are tailored for each host video frame. Indeed, the frame index is firstly binary modulated and then, serves as a primary watermark. Subsequently, a second content-based authentication watermark is built up using frame blocks texture features. To improve the security aspect, Torus automorphism mapping is applied under the second watermark before being embedded. The suggested scheme operates in the hybrid domain. In fact, the primary watermark is infused in the host frame following a spatial domain-based embedding technique. However, a frequency domain-based watermarking method, which combines Lifting Wavelet Transform (LWT) and Singular Value Decomposition (SVD), is involved to hide the second watermark. Herein, the most textured blocks are chosen as second watermark holders to enhance the watermarked video perceptual quality. During the detection process, the dissimilarity between the second extracted watermark and its reconstructed version as well as the mismatch between the observed index and the extracted one enable to reveal spatial and temporal tampering, respectively. Experimental results show that the proposed scheme achieves a good visual quality with a large embedding capacity. Furthermore, it can efficiently withstand non-malicious processing while being fragile to frames manipulations and content modification attacks. Moreover, it ensures an accurate tampered areas localization. [ABSTRACT FROM AUTHOR]

This paper investigates the stability behaviour of externally interfered discrete-time system under the influence of various finite-register length nonlinearities. Such nonlinearities include quantization and overflow which are commonly originated during hardware implementation. The proposed criteria guarantee the realization of limit-cycle free discrete-time system in the presence of external disturbances and finite-register length nonlinearities. The stability conditions are fabricated by employing passivity-based approach under combinations of quantization and overflow nonlinearities. With the help of the appropriate Lyapunov function along with the passivity condition, the asymptotic stability for the discrete-time system is also ensured under zero disturbance. Relevant examples are presented to show the applicational scope of the work. [ABSTRACT FROM AUTHOR]

This paper is concerned with the design of state-feedback guaranteed cost controller (GCC) for linear two-dimensional (2-D) uncertain discrete delayed systems with actuator saturation (AS). The 2-D system under consideration is represented by the Roesser model. The linear matrix inequality-based conditions for the design of GCC are developed. The proposed method not only ensures that closed-loop system trajectories converge to the origin, but it also provides a satisfactory performance level under all permissible system uncertainties. A convex optimization problem is also formulated for the design of optimal GCC. The GCC design problem for 2-D systems with AS and no delay is also examined. The problem of GCC design for 2-D systems in absence of AS and state delay is also discussed. Several examples are given to illustrate the applicability of the presented results. [ABSTRACT FROM AUTHOR]

Invisibility, robustness and payload are three indispensable and contradictory properties for any image watermarking systems. To achieve the tradeoff among the three requirements, statistical watermarking approaches have received increasing attention in recent years. But, most existing schemes often bear a number of drawbacks, in particular: (1) They mainly utilize transform coefficients, which are always fragile to some attacks, especially global geometric transforms, for watermark inserting and statistical modeling; (2) the adopted statistical models always cannot capture accurately both marginal distribution and various strong dependencies between coefficients; and (3) the used parameter estimation usually has high time complexity and poor computational accuracy. This has motivated us to introduce in this paper a novel statistical image watermarking in singular value decomposition (SVD)-undecimated discrete wavelet transform (UDWT) difference domain using vector Alpha Skew Gaussian (VB-ASG) distribution. We begin with a detailed study on the robustness and statistical characteristics of local SVD-UDWT difference coefficients of natural images. This study reveals the excellent robustness, highly non-Gaussian marginal statistics and strong dependencies of local SVD-UDWT difference coefficients. We also find that conditioned on their generalized neighborhoods, the local SVD-UDWT difference coefficients can be approximately modeled as vector Alpha Skew Gaussian (VB-ASG) variables. Meanwhile, model parameters can be estimated effectively by using approximate maximum likelihood estimation (AMLE) approach. Based on these findings, we model local SVD-UDWT difference coefficients using VB-ASG model that can capture marginal statistics and strong dependencies. Finally, we develop a new statistical image watermark decoder using the VB-ASG model and maximum likelihood (ML) decision rule. Our experimental evaluation results validate that our image watermarking leads to performance improvements comparable to several state-of-the-art statistical watermarking methods and some approaches based on convolutional neural networks. In particular, under the condition of the same watermarking capacity, the imperceptibility and robustness of this method show certain superiority compared with other algorithms. [ABSTRACT FROM AUTHOR]

In this paper, a novel adaptive genetic algorithm (AGA) is presented for the optimization goals. The presented AGA is discussed and implemented on a power amplifier (PA), which is designed for 24 GHz and 5G applications and fabricated in a 65-nm CMOS process. The PA is optimized by the AGA for the high power-added efficiency (PAE), i.e., optimum. In the AGA, we aimed to avoid the local optima and slow convergence rate that exist in the conventional genetic algorithm (CGA). In the AGA, we proposed a parameter tuning method to fine-tune the set of PA circuit component values for faster optimization of the PA to have a high PAE. The proposed AGA provides a significant speed in optimization process as compared to the CGA, and the execution time of the AGA is faster than that of the CGA. The proposed AGA is also verified, and its performance is compared with that of the CGA through multiple multidimensional mathematical benchmark functions. The PA performance parameters are measured, and the results showed that the optimized PA achieves a high gain of 29.9 dB. The Psat of PA is measured as 14.21 dBm, and IIP3 is 13.8 dBm. The simulated result shows the optimized PA with the AGA has PAE of 49.7%, while that with the CGA has PAE of 47.5% at 24 GHz. The final measured PAE of AGA's optimized PA is 47.1%. The chip area of PA is 0.29 mm2. [ABSTRACT FROM AUTHOR]

In recent years, long short-term memory (LSTM) networks have been extensively used in various domains such as machine interpretation, language displaying, and so forth. The paper proposes two architectures for accelerating matrix vector multiplication (MVM) in the LSTM network. The first proposal introduces an architecture for the MVM unit by incorporating the concept of Activity Span Reduction (ASR). Here, the partial products for inner product computations (IPC) are generated using ASR logic. This logic ensures that the computational units are active for an optimal duration, effectively utilizing time without compromising the functionality. This approach reduces power consumption as well as area requirements. The ASR technique is able to achieve a 65.9% reduction in power for MVM and occupies 43.77% fewer logic cell units when compared to the state of the art. The second proposal emphasizes the development of a two-stage recursive MVM architecture for the LSTM network. A partial product generator unit using recursive arithmetic architecture is proposed for the IPC in MVM, intended to reduce the power and area consumption. This MVM architecture provides a 6.8% improvement in the maximum clock frequency without an increase in power consumption. The utilization of the proposed designs in the development of the LSTM architecture provides additional evidence supporting the effectiveness of the proposed MVMs. [ABSTRACT FROM AUTHOR]

Frequency domain filtered-x least mean square algorithms can reduce the computational complexity of the time domain counterpart with long filters; however, they suffer from large block delay, additional quantization error due to large size transformations and implementation difficulties in existing DSP hardware. In this paper, a time–frequency domain flexible structure is proposed using the partitioned block frequency domain adaptive filtering technique, which has no signal path delay and is well suited for low-cost DSP implementation. The proposed structure divides the long filters into many equal partitions and carries out the control filter update in frequency domain while generating the control signal in both time and frequency domains, thereby eliminating the forward path delay completely while maintaining low computational complexity. The proposed structure has a potential benefit for controlling broadband noise, where the causality constraint is more important. The simulation results using the measured acoustic paths demonstrate that the proposed structure maintains similar control performance as that of the time domain algorithm but with much less computational complexity. Furthermore, the tracking performance of the proposed structure under different levels of measurement noise is investigated. [ABSTRACT FROM AUTHOR]

A novel delay cell circuit for differential ring oscillator (DRO) with large tuning range along with application in charge pump phase lock loop (CP-PLL) frequency synthesizer has been presented in this paper. Using 0.18 μ m CMOS technology with power supply of 1.8 V, the two DRO architectures: 3-stage and 5-stage, were built and simulated. In both 3-stage and 5-stage DROs, single controlled voltage is employed. The suggested 3-stage and 5-stage DRO circuits generate a tuning range of 96.77 MHz - 5.296 GHz and 36.33 MHz - 2.803 GHz, respectively. The % total harmonic distortion (%THD) of both DRO architectures is also evaluated. The suggested 3-stage and 5-stage DROs consume 6.63 mW and 11.05 mW power at an oscillation frequency of 4.76 GHz and 2.479 GHz, respectively. At an offset frequency of 10 MHz from the oscillation frequency, the proposed circuits have phase noise of - 119.93 dBc/Hz and - 128.24 dBc/Hz, respectively. The layout of proposed design has been drawn and pre- and post-layout simulation results show satisfactory variations of tuning range and phase noise of proposed design. The suggested circuit's robustness is verified with the help of PVT and Monte Carlo analysis. When compared to contemporary research, the proposed DROs have the widest tuning range. Proposed DRO application in CP-PLL frequency synthesizer has locking time of 1.17 μ s and shows good settling behaviour with dynamic parameter variations. [ABSTRACT FROM AUTHOR]

In this paper, we investigate the finite-time output feedback control for discrete asynchronous switched systems with input saturation and output saturation based on MDADT. Firstly, we solve the input saturation and output saturation problem by employing a low-gain output feedback controller and the sector nonlinear condition. Secondly, we consider the asynchronism under event-triggered strategy and use the MDADT approach for finite-time stabilization and boundedness of the discrete-time switched systems under asynchronous switching. Some sufficient conditions in terms of LMIs and the appropriate MDADT are provided. Finally, two examples are given to show the effectiveness of the theoretical results. [ABSTRACT FROM AUTHOR]

A minimum-disturbance description of adaptive beamforming algorithms opens up the derivation of novel methods for linearly constrained settings. Under this context, this paper proposes two algorithms, the IPCNLMS and the IPCNLMS- ℓ 0 , and discusses their application for adaptive beamforming using a uniform rectangular array. These algorithms combine both proportionate and norm constraint strategies that are inserted in the traditional CNLMS algorithm in a smooth and rigorous way. The idea of jointly implementing CNLMS and IPNLMS algorithms is used to achieve faster convergence and the ability to attenuate interfering signals from multiple directions while using adaptive beamforming. The IPCNLMS updates each filter coefficient independently by adjusting the adaptation step size proportionally to the magnitude of the estimated filter coefficient. It is based on the ℓ 1 -norm penalty to exploit the convergence speed of the system. To take this idea further, the ℓ 0 -norm penalty is also considered in the IPCNLMS- ℓ 0 proposition. Simulations demonstrate that the proposed algorithms present faster convergence, under equivalent conditions of asymptotic performance, even when more realistic coupling effects between the array elements are taken into account. Beam pattern results show that the proposed algorithms are capable of achieving the optimum solution from LCMV as well as the CNLMS. [ABSTRACT FROM AUTHOR]

The cry of a premature infant is an attempt to connect with its mother or others. The newborns are communicated in different ways depending on the reason for their screams. In recent days, the preprocessing, feature extraction, and classification of audio signals require expert attention and a lot of effort. In this paper, a novel deep convolutional recurrent neural network (DCR net) has been proposed to classify the premature infant cry signal into different categories. The acquisition of the cry signal generally requires a lengthy observation period and several activity processes to obtain all the signals of the premature infant. The relevant multi-class frequency features are extracted by using the MFCC (Mel-frequency cepstral coefficient), BFCC (Bark-frequency cepstral coefficient), and LPCC (linear prediction cepstral coefficient) features, which are combined to create a fused feature matrix that is helpful in the classification of pathological crying. Based on these features, the DCR net is used to classify sounds in the premature infant cry. The sound of the target cry signal is classified into five categories: "neh" means hunger, "heh" means discomfort, "eh" means burping, "eair" means cramps, and "owh" means fatigue. The efficiency of the DCR net was estimated with some metrics such as specificity, precision, accuracy, recall, and F1 score. The experimental fallouts disclose that the proposed DCR net attains a better classification accuracy of 97.27% for identifying infant cry signals. The DCR net increases the overall performance range by 8.61%, 11.58%, 0.54%, and 17.03% better than SVM-RBF, MFCC-SVM, optimized deep learning model, and hidden Markov model, respectively. [ABSTRACT FROM AUTHOR]

Speech enhancement is an important method for improving speech quality and intelligibility in noisy environments. An effective speech enhancement model depends on precise modelling of the long-range dependencies of noisy speech. Several recent studies have examined ways to enhance speech by capturing the long-term contextual information. For speech enhancement, the time–frequency (T–F) distribution of speech spectral components is also important, but is usually ignored in these studies. The multi-stage learning method is an effective way to integrate various deep learning modules at the same time. The benefit of multi-stage training is that the optimization target can be iteratively updated stage by stage. In this paper, speech enhancement is investigated by multi-stage learning using a multi-stage structure in which time–frequency attention (TFA) blocks are followed by stacks of squeezed temporal convolutional networks (S-TCN) with exponentially increasing dilation rates. To reinject original information into later stages, a feature fusion block (FB) is inserted at the input of later stages to reduce the possibility of speech information being lost in the early stages. The S-TCN blocks are responsible for temporal sequence modelling tasks. The time–frequency attention (TFA) is a simple but effective network module that explicitly exploits position information to generate a 2D attention map to characterize the salient T–F distribution of speech by using two branches, time-frame attention and frequency attention in parallel. Extensive experiments have demonstrated that the proposed model consistently improves the performance over existing baselines across two widely used objective metrics such as PESQ and STOI. A significant improvement in system robustness to noise is also shown by our evaluation results using the TFA module. [ABSTRACT FROM AUTHOR]

This paper probes into the problem of predefined-time consensus tracking for a class of second-order multi-agent systems (MASs) subject to external disturbances based on a predefined-time observer. A predefined-time control strategy based on the use of terminal sliding mode control method is proposed, which can achieve consensus tracking in a predefined time. Firstly, the observer is designed so that observed states and leader states reach consensus. Secondly, according to the designed controller, states of each agent and observed states achieve consensus, namely, leader-following consensus is realized. By employing Lyapunov stability theory, some sufficient criteria are obtained to achieve predefined-time consensus tracking for second-order MASs with disturbances. Finally, some numerical examples are presented to prove the validity of our designs. [ABSTRACT FROM AUTHOR]

The complex harmonic wavelet (CHW) for the discrete signal is orthogonal, compact support in the frequency domain but is not complex conjugate symmetric for positive and negative half-planes. The modified complex harmonic wavelet (MCHW) transform is the improved version of CHW as it is the complex conjugate symmetry along with other desired properties of CHW such as orthogonality, and compact support in the frequency domain. Due to the complex conjugate symmetry, MCHW has lesser computational complexity compared to CHW. This paper introduces a new VLSI architecture for MCHW for hardware implementation and prototyped on a commercially available virtex5 field-programmable gate array (FPGA). For the validation of the proposed implementation, the real-time captured results in the logic analyzer are verified with simulation results. The maximum operating frequency targeting the above-mentioned FPGA device is reported as 92.82 MHz. The total on-chip power of the above implementation is 1.117W, out of which 84 mW is the dynamic power dissipation at a toggle rate of 12.5 %. Finally, for the area utilization of the above implementation, its resource utilization targeting the above FPGA device is reported. [ABSTRACT FROM AUTHOR]

In this work, we propose the unfolded versions of the coprime transformed nested array (CTNA), which exploits the sum difference coarray (SDCA). We know that the CTNA has attractive properties in achieving more degrees of freedom (DOF) in direction-of-arrival estimation. However, the mutual coupling of the CTNA cannot be neglected. Here, we present the unfolded versions of the CNTA, specifically UCTNA-I (Unfolded CTNA) and UCTNA-II, which are obtained by flipping two different subarrays of the CTNA. The proposed configurations significantly reduce the mutual coupling of the CTNA without compromising the DOF it achieves. Vectorized conjugate augmented MUSIC algorithm is used to explore the SDCA of the array. The detailed properties of the arrays with the weight functions are provided in the paper. The optimization of the arrays is also carried out. Simulations are performed to prove the superiority of the proposed arrays by comparing them with the other existing arrays. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new filtering algorithm, the variational Bayesian-based adaptive maximum correntropy generalized high-degree cubature Kalman filter, which is designed to improve filtering accuracy under conditions of unknown measurement noise covariance and measurement outliers. Considering that the generalized high-degree cubature rule can solve the high dimensional nonlinear problem well, based on the generalized high-degree cubature Kalman filter, the variational Bayesian method is utilized to approximate the measurement noise covariance, and the maximum correntropy criterion is used to reduce the influence of the measurement outliers on the state estimation. Additionally, we introduce a resistance factor to correct measurement values and optimize the kernel bandwidth for different types of noise. Simulation experiments on target tracking and integrated navigation demonstrate that our proposed algorithm effectively suppresses unknown time-varying noise and non-Gaussian mutation noise, outperforming existing filtering algorithms in terms of estimation accuracy, robustness, and adaptability. [ABSTRACT FROM AUTHOR]