The performance of personal sound systems is often degraded by inaccurate acoustic measurements. To achieve robust control while balancing acoustic contrast and signal distortion, this work proposes a robust hybrid optimization method that exploits both acoustic contrast control and pressure matching (ACC-PM). The method addresses perturbations caused by uncertainties in the acoustic transfer functions such as temperature changes, head movement, etc, modeled as norm-bounded uncertainties. Although the resulting worst-case optimization is inherently non-convex, it is reformulated as a second-order cone programming problem, which can be efficiently solved. Numerical simulations demonstrate the effectiveness of the proposed robust ACC-PM algorithm, showing an improvement over 18% in terms of AC compared to vanilla ACC-PM.

Modulo analog-to-digital converters (ADCs) offer a potential solution to the clipping challenges in conventional ADCs by folding signals that exceed the threshold. This makes them suitable for high-amplitude signal acquisition in wide dynamic range applications. In this demonstration, we present a 2-channel modulo ADC system implemented on a field-programmable gate array (FPGA) with an integrated real-time recovery algorithm. By managing both signal folding and recovery entirely in hard-ware, the system ensures low-latency processing. The FPGA efficiently handles high-bandwidth signals, making it a promising option for applications that require robust performance in high-amplitude signal environments.

In this paper, an efficient hybrid beamforming scheme is proposed for the relay node in mmWave relay systems based on deep unfolding. The optimal digital beamformer at the relay node is decomposed into an analog-digital- analog (A-D-A) hybrid structure, which can be formulated as a matrix factorization problem. The analog beamforming matrices and digital beamforming matrix in the A-D-A hybrid structure are efficiently solved by a deep unfolding neural network, which is unfolded based on the projected gradient descent algorithm. Numerical results verify that performance of the proposed scheme is close to the fully-digital solution, which prevails traditional model-based algorithms with a fast runtime.

In multi-user adversarial scenarios involving external malicious jamming and internal co-channel interference, environmental instability and increased decision-making dimensions cause traditional deep reinforcement learning (DRL)-based anti-jamming schemes to suffer from insufficient exploration. Agents must choose policies from a large action set, leading to a significant decline in anti-jamming performance. To address these issues, this paper proposes a multi-agent discrete soft actorcritic (MA-DSAC) algorithm-based collaborative anti-jamming strategy, integrating frequency, power, and modulation-coding domains. This strategy first introduces a Markov game to model and analyze the multi-user anti-jamming problem. Next, the soft actor-critic (SAC) algorithm is discretized to handle the multi-dimensional discrete action space. Finally, through information exchange between communication transceivers and based on a centralized training with decentralized execution (CTDE) framework, it is extended to a multi-agent DRL algorithm to achieve efficient multi-user cooperative anti-jamming. Simulation results show that in various anti-jamming scenarios with both fixed-mode and intelligent jammers, the proposed anti-jamming strategy’s performance improves by more than 25% compared to traditional value-based DRL strategies, including independent deep Q-network (I-DQN) and multi-agent virtual exploration in deep Q-learning (MA-VEDQL). Furthermore, through information exchange between communication transceivers, the instability problem of multi-agent DRL is effectively alleviated, enabling the communication transceivers to balance competition and cooperation. Consequently, its anti-jamming performance improves by more than 6% compared to the independent DSAC (I-DSAC) strategy.

This paper explores noise-mismatched models using the Hellinger distance. In many applications, the design/training stage often assumes an independent and identically distributed (i.i.d.) Gaussian prior noise, but the real world introduces Gaussian noise with arbitrary covariance, creating a mismatch. We analyze the impact on system output and study optimal injected noise intensity for training/design. While theory assumes Gaussian sources, it provides guidance for non-Gaussian settings too. Experiments with Cycle-GAN for image-to-image translation validate the theory, producing results consistenting with derivations. Overall, this work provides theoretical and empirical insights into designing systems robust to noise uncertainties beyond simplified assumptions.

This paper presents a theoretical analysis of multi-channel modulo analog-to-digital converters (ADCs) for high-dynamic range sampling under bounded noise. In particular, we derive the maximum error tolerance in terms of ADC dynamic range, signal dynamic range, and channel number. Additionally, we present closed-form expressions for ADC thresholds, ensuring near-optimal error resilience, and analyzing the minimal bit-depth needed for stable recovery. Compared to single-channel modulo ADCs, our approach achieves superior error tolerance with reduced sampling rates. Moreover, it demands a minor bit rate increase compared to conventional ADCs but operates with a significantly smaller ADC dynamic range.

Image-to-image (I2I) translation is vital in computer vision tasks like style transfer and domain adaptation. While recent advances in GAN have enabled high-quality sample generation, real-world challenges such as noise and distortion remain significant obstacles. Although Gaussian noise injection during training has been utilized, its theoretical underpinnings have been unclear. This work provides a robust theoretical framework elucidating the role of Gaussian noise injection in I2I translation models. We address critical questions on the influence of noise variance on distribution divergence, resilience to unseen noise types, and optimal noise intensity selection. Our contributions include connecting $f$-divergence and score matching, unveiling insights into the impact of Gaussian noise on aligning probability distributions, and demonstrating generalized robustness implications. We also explore choosing an optimal training noise level for consistent performance in noisy environments. Extensive experiments validate our theoretical findings, showing substantial improvements over various I2I baseline models in noisy settings. Our research rigorously grounds Gaussian noise injection for I2I translation, offering a sophisticated theoretical understanding beyond heuristic applications.

This paper introduces five novel deep-learning architectures for speech enhancement. Existing methods typically use time-domain, time-frequency representations, or a hybrid approach. Recognizing the unique contributions of each domain to feature extraction and model design, this study investigates the integration of waveform and complex spectrogram models through cross-domain fusion to enhance speech feature learning and noise reduction, thereby improving speech quality. We examine both cascading and parallel configurations of waveform and complex spectrogram models to assess their effectiveness in speech enhancement. Additionally, we employ an orthogonal projection-based error decomposition technique and manage the inputs of individual sub-models to analyze factors affecting speech quality. The network is trained by optimizing three specific loss functions applied across all sub-models. Our experiments, using the DNS Challenge (ICASSP 2021) dataset, reveal that the proposed models surpass existing benchmarks in speech enhancement, offering superior speech quality and intelligibility. These results highlight the efficacy of our cross-domain fusion strategy.

Annual Conference of the International Speech Communication Association

With the developments of various multi-function communication services, acoustic echoes and background noises inevitably appear in hands-free calling occasions. Different from using the combination of neural network and traditional acoustic echo cancellation (AEC) method, this paper directly proposes a time-frequency complex cascaded neural network (TFCN) for echo cancellation and noise suppression. To that aim, in frequency domain, complex LSTM layers are employed to process the real and imaginary signals. After that, an end-to-end time domain network is designed using dilated convolution layers to further remove residual interferences. By adding rich delay information to the dataset and optimizing the model by a weighted loss function, the generalization ability of the model is also improved. The extensive experimental results show that the proposed frame-work is robust to blind test datasets, effectively removes echoes and noises, and achieves an excellent performance on AECMOS scores. The subjective mean score of the proposed method is 4.37, which is 0.50 higher than the INTERSPEECH2021 AEC-Challenge baseline.

In this paper, robust hybrid beamforming designs are developed for multi-user millimeter-wave multiple-input multiple-output relay systems in the presence of correlated channel state information errors. Analog beamforming matrices are designed to maximize the equivalent channel gains. Baseband beamforming matrices are solved by optimizing the upper bound of the averaged achievable sum rate. Simulation results show that our algorithm achieves better performance compared with previous existing hybrid beamforming designs.

Inverse synthetic aperture radar (ISAR) imaging for maneuvering targets has always been a challenging task due to azimuth time-varying Doppler frequency modulation, especially under moving platform condition. In this case, the common assumption that the image projection plane (IPP) of the radar line-of-sight (LOS) direction is constant during coherent processing interval (CPI) is invalid. To address this issue, a novel ISAR imaging approach for maneuvering targets is proposed by exploiting nonstationary IPP in this article. First, considering time-varying LOS direction, the new geometric and signal models are developed, where 2-D spatial-variant phase error is mainly deduced. After that, a parametric image entropy minimum optimization combined with efficient particle swarm optimization (PSO) is used to obtain optimal motion parameters. In doing so, 2-D spatial-variant phase error terms are compensated accurately to produce well-focused ISAR image. Finally, the effectiveness and superiority of the proposed algorithm are verified by the simulation results and electromagnetic scattering data.

The front-wall clutter removal in the case of the through-the-wall radar (TWR) system is studied in this work. To remove the wall clutter, its low-rank property is utilized, and at the same time, the sparse property of the target returns is exploited to perform target reconstruction. To account for the unparalleled setting of the antenna and the wall, a weighted nuclear norm minimization (WNNM) is employed, and the resulting problem is solved in an alternating manner. In addition, different transmitted waveform signals, including monofrequency and stepped-frequency waveforms, are used to demonstrate their effects on the clutter suppression performances. The experimental results show that the proposed WNNM with stepped-frequency waveform outperforms other approaches.

Since only the spatial feature information of ship target is utilized, the current deep learning-based synthetic aperture radar (SAR) ship detection approaches cannot achieve a satisfactory performance, especially in the case of multiscale or rotations, and the complex background. To overcome these issues, a novel multidimensional domain deep learning network for SAR ship detection is developed in this work to exploit the spatial and frequency-domain complementary features. The proposed method consists of the following main three steps. First, to learn hierarchical spatial features, the feature pyramid network (FPN) is adopted to produce ship target spatial multiscale characteristics with a top-down structure. Second, with a polar Fourier transform, the rotation-invariant features of SAR ship targets are obtained in the frequency domain. After that, a novel spatial-frequency characteristics fusion network is then presented, which seeks to learn more compact feature representations across different domains by updating the parameters of sub-networks interactively. The detection results are obtained due to utilizing the multidimensional domain information, and we evaluate the effectiveness of the proposed method using the existing SAR ship detection data set (SSDD). The results of the proposed method outperform other convolutional neural network (CNN)-based algorithms, especially for multiscale and rotation ship targets under complex backgrounds.

Owing to the highly maneuvering character involved in targets, the nonuniform 3-D rotation motions make the assumption that the image projection plane (IPP) is constant during coherent processing interval (CPI) invalid. In this work, an efficient approach is proposed in ISAR imaging for highly maneuvering targets with nonstationary IPP. First, to reasonably describe the mobility of a highly maneuvering motion target, the geometry and signal model with nonstationary IPP are established, where the high-order phase model is deduced to describe the 2-D spatial-variant phase errors. Second, based on the developed signal model, considering the cost function obtained via conventional image entropy with local extremum, the subarray averaging operation in conjunction with entropy is utilized to accelerate the global optimal convergence. Finally, the accurate 2-D spatial-variant phase errors compensation terms are generated to produce the well-focused ISAR images. Compared with existing methods, the main advantages of this work are: 1) the geometry and signal model of the target with nonstationary IPP are established; 2) the subarray averaging operation in conjunction with image entropy is utilized to accelerate the global optimal convergence; and 3) the high-order signal model is derived to present the 2-D spatial-variant phase errors. Several numerical experiments using simulated data and electromagnetic data are conducted to demonstrate the validity of the proposed algorithm and signal model.

Convolutional neural networks (CNNs) have emerged as powerful tools for image retrieval and classification. In this paper, we study binary embedding of CNN-based image descriptors for resource-constrained devices. We propose two classes of fast computable and memory-efficient dimension reduction operators using Golay-Hadamard matrices (GHMs), which are constructed by multiplying the columns of a Hadamard matrix with a Golay sequence. Simulation results on CNN-based instance image retrieval and classification show that GHM-based operators can offer competitive performance to those of full random Gaussian matrices and Gaussian circulant matrices at much lower computational cost and storage space. This implies the potential of proposed approaches on devices with low memory, bandwidth and power restrictions.

Recently, there is an increased interest in the study of modulo analog to digital converters (ADCs). These new systems can reconstruct a signal whose amplitude is much higher than the conventional ADC's dynamic range. Modulo ADCs are characterized by their modulo threshold and in the current literature, all existing works are limited to real-valued moduli. In this paper, we propose multi-channel modulo samplers with complex-valued moduli to sample a band-limited complex signal. Specifically, we discuss the construction of complex divisors from Gaussian integers and propose their efficient implementations. A memory-efficient, closed-form recovery algorithm is also proposed. Simulation results demonstrate that the proposed systems can provide stable reconstruction of a high dynamic range complex-valued signal at low sampling rates.

This article addresses the problem of tracking targets behind the wall using through-the-wall radar. To that end, the wall reflection, i.e., clutter, must be eliminated first because it interferes with the subsequent image formation operation. The low-rank of the clutter and sparseness of the useful signal are utilized to devise a joint low-rank and sparse framework to simultaneously suppress the clutter and recover the target returns, where alternating direction method of multipliers (ADMM) approach is developed to solve the corresponding optimization. Since then, an effective observation window scheme is proposed to locate the target and further to facilitate the tracking process. The tracking is finally provided by Kalman filter and particle filter. The numerical studies are provided to demonstrate that the performance of the proposed framework is superior to that of other methods in terms of clutter removal and tracking accuracy.

Owing to the large range cell migration (RCM) and fast time-variant Doppler frequency modulation (DFM) generated by rapidly spinning targets, it is difficult to efficiently obtain well-focused inverse synthetic aperture radar (ISAR) images via conventional algorithms because of the multidimensional search requirement. Inspired by the inherent azimuth spatial invariance in strip-map synthetic aperture radar (SAR) imaging mode, an efficient range-Doppler domain ISAR imaging method for rapidly spinning targets is proposed in this article. First, echo signal is transformed into range-Doppler domain and its precise analytical expression is derived according to the principle of stationary phase (POSP). Second, the energy of scatterers distributed in different range cells is extracted along the rotating radius. By doing so, the energy is concentrated in the same range cell. After that, the high-order phase terms of the signal are compensated and the CLEAN technique is also applied to reduce the sidelobes of a strong scatterer. Finally, 3-D ISAR image of the spinning target is reconstructed by projecting the spatial parameters to 3-D cylindrical coordinates. Furthermore, in this article, the output signal-to-noise ratio (SNR) gain, anti-noise performance, the mismatched phase error, and the computational complexity analyses are also provided. Compared with existing approaches, the proposed method has advantages in the computational complexity and low SNR environment thanks to the only 1-D search and the coherent integration gain obtained. Both the theoretical derivations and the simulated results demonstrate the effectiveness of the proposed method.

The recovery problem of synthetic aperture radar (SAR) signal in the presence of radio frequency interference (RFI) is studied. To perform RFI suppression, in this paper, the RFI is modeled as the combination of multiple complex sinusoids such that the RFI suppression problem becomes a frequency estimation one. To accurately estimate model parameters, by exploiting sparse representation of the RFI, a gridless approach based on atomic norm minimization is proposed, which completely removes the off-grid issue. Finally, to recover the SAR signal, a joint scheme is devised to simultaneously perform the RFI suppression and the SAR signal recovery under an optimization framework. The resultant optimization is efficiently solved by a two-step process based on the coordinate descent approach. Simulation results and real-world experiments are provided to show the superior performance of the proposed approach.

In this paper, two new decoding algorithms to decode Reed-Solomon codes during transmission over burst Rayleigh fading channels with additive white Gaussian noise (AWGN) are proposed. They only conduct error correction for coded symbols located in the pure AWGN region and conduct error and erasure correction for those symbols located in the burst fading region by treating those coded symbols that are very likely erroneous as erasures. The first algorithm does not need to know the fading locations in advance, while the second algorithm assumes that the fading locations are known. In addition, the performance of such two algorithms is studied when a pre-computed threshold is used to determine the erasures of the code. Simulation results show that our proposed algorithms not only significantly perform better than the classic Berlekamp-Messay algorithm with a comparable computational complexity but also achieve a better tradeoff between the performance and the computational complexity when compared with other existing algorithms. In particular, our algorithms exhibit excellent robustness for tested various code parameters and fading configurations. Furthermore, a more detailed mathematical analysis is also developed in this paper in order to estimate the performance of the new algorithms in the burst Rayleigh fading channels. We observe that the performance of the first algorithm can only be estimated relatively accurately when encountering burst deep-fading, whereas the performance prediction for the second algorithm is always in agreement with the simulation results for various fading cases.

To better interpret the inverse synthetic aperture radar (ISAR) imaging results, it is highly desirable to present them in the homogeneous range-cross-range domain, rather than the conventional range-Doppler (RD) domain. This process is referred to as cross-range scaling and the rotating angle velocity (RAV) of the moving target must be estimated first to achieve that goal. In this paper, an efficient cross-range scaling approach based on 2-D discrete wavelet transform (2D-DWT) and pseudopolar fast Fourier transform (PPFFT) is developed. To be exact, first, 2D-DWT is applied to two sequential ISAR images to obtain the dominant feature points based on the fact that the ISAR images are usually redundant for estimating RAV. By doing so, the data dimensional reduction and noise suppression are also realized. After that, second, via the efficient PPFFT, two sequential RD ISAR images are mapped into the pseudopolar coordinate to convert the rotational motion into the translational motion along the pseudo angle direction. Finally, to estimate the RAV, a new normalized correlation cost function is constructed and the Golden section algorithm is employed to efficiently find the optimal RAV. Compared with the conventional methods, the advantages of the proposed method are threefold: 1) the rotation center of a target is no longer required prior; 2) without the interpolation operation and the utilization of data dimensional reduction via 2D-DWT, the computational complexity of the proposed method is significantly reduced;and 3) the accurate RAV estimation is achieved in the case of low signal-to-noise ratio condition. The results from both the simulated and the measured data demonstrate that the proposed approach outperforms the state-of-the-art algorithms in terms of the estimation accuracy and computational complexity.

This paper addresses the problem of recovering a synthetic aperture radar (SAR) signal that is corrupted by both radio frequency interference (RFI) and wideband interference (WBI). The time–frequency domain is utilized for both the SAR signal and interference in the form of sparse representations. By doing so, a unified framework that allows one to suppress both the RFI and WBI while recovering the SAR signal can be developed. The resulting framework is an optimization problem that is efficiently solved using a customized alternating direction method of multipliers approach. Finally, simulation results are provided to demonstrate that the performance of the joint estimation algorithm is superior to the performances of other methods in terms of both subjective and objective evaluation standards.

In this paper, an efficient decoding algorithm is developed to facilitate faster decoding of the binary systematic quadratic residue (QR) codes. It is based on the difference of syndromes (DS), and hence, is called the DS algorithm hereinafter. This new method combines the advantages of the syndrome-weight algorithm and properties of the cyclic codes. Actually, it is a natural generalization of the cyclic weight (CW) algorithm for the (47, 24, 11) QR code developed by Lin et al., and its validity of decoding any binary systematic QR code is also proved. The complexity analysis and simulation results show that the DS algorithm dramatically reduces the decoding complexity without performance loss and considerably requires less memory when compared with previous ones. Utilizing the (47, 24, 11), the (71, 36, 11), the (73, 37, 13), and the (89, 45, 17) QR codes as examples, the DS algorithm not only significantly improves the decoding efficiency, but also saves the memory evidently. When the (23, 12, 7) QR code is considered, the DS algorithm performs almost as well as the currently known best algorithm in terms of decoding efficiency and memory requirements. Thus, all QR codes of lengths less than 100 can be decoded efficiently by using the proposed algorithm. Especially, when the proposed algorithm is applied to decode the (89, 45, 17) QR code, the best one among all the QR codes of lengths less than 100, the decoding speed is raised 26 times and the memory is saved up to 76.6% in comparison with the existing fastest decoding algorithm.

The detection and parameters estimation of linear frequency-modulated (LFM) signal are important for modern radar applications, but they are also challenged by the fact that echo signal is often of low signal-to-noise ratio (SNR) due to reasons of long imaging distance and/or limited transmitted power, and the target of small size and/or hidden characteristics. To enhance the SNR, in our previous work, a novel coherently integrated cubic phase function (CICPF) was recently developed for the parameters estimation of the multicomponent LFM signal. In the CICPF, the auto-terms are coherently integrated to enhance the performance in the case of low SNR and also to suppress the cross-terms and spurious peaks. In this paper, as an extension of our previous work, the theoretical performance analyses including several important properties and the fast implementation are provided. Furthermore, the asymptotic mean squared error of a CICPF-based estimator as well as the output SNR of a CICPF-based detector are theoretically derived in closed-forms. From the performance point of view, the proposed CICPF attains the Cramer-Rao bound at low input SNR. The complexity analysis also indicates that the CICPF with the nonuniform fast Fourier transform is computationally efficient without needing the interpolation operation and parameter search. Numerical studies of the CICPF confirm the theoretical analysis and demonstrate superior performance of the proposed approach compared with other state-of-the-art approaches, especially under the low-SNR condition. Finally, the proposed CICPF is applied for the ground moving target imaging in synthetic aperture radar. Results using simulated and experimental data demonstrate that it provides an effective means to obtain well-focused image for ground moving targets.

In this paper1, a method to search the subsets I and J needed in computing the unknown syndromes for the (73, 37, 13) quadratic residue (QR) code is proposed. According to the resulting I and J, one computes the unknown syndromes, and thus finds the corresponding error-locator polynomial by using an inverse-free BM algorithm. Based on the modified Chase-II algorithm, the performance of soft-decision decoding for the (73, 37, 13) QR code is given. This result is never seen in the literature, to our knowledge. Moreover, the error-rate performance of linear programming (LP) decoding for the (73, 37, 13) QR code is also investigated, and LP-based decoding is shown to be significantly superior in performance to the algebraic soft-decision decoding while requiring almost the same computational complexity.

In this paper, a method to search the set of syndromes' indices needed in computing the unknown syndromes for the (73, 37, 13) quadratic residue (QR) code is proposed. According to the resulting index sets, one computes the unknown syndromes and thus finds the corresponding error-locator polynomial by using an inverse-free Berlekamp-Massey (BM) algorithm. Based on the modified Chase-II algorithm, the performance of soft-decision decoding for the (73, 37, 13) QR code is given. This result is new. Moreover, the error-rate performance of linear programming (LP) decoding for the (73, 37, 13) QR code is also investigated, and LP-based decoding is shown to be significantly superior in performance to the algebraic soft-decision decoding while requiring almost the same computational complexity. In fact, the algebraic hard-decision and soft-decision decoding of the (89, 45, 17) QR code outperforms that of the (73, 37, 13) QR code because the former has a larger minimal distance. However, experimental results indicate that the (73, 37, 13) QR code outperforms the (89, 45, 17) QR code with much fewer arithmetic operations when using the LP-based decoding algorithms. The pseudocodewords analysis partially explains this seemingly strange phenomenon.