Pub Date : 2024-12-03DOI: 10.1109/TSP.2024.3510680
Ricardo A. Borsoi;Konstantin Usevich;David Brie;Tülay Adali
Coupled tensor decompositions (CTDs) perform data fusion by linking factors from different datasets. Although many CTDs have been already proposed, current works do not address important challenges of data fusion, where: 1) the datasets are often heterogeneous, constituting different “views” of a given phenomena (multimodality); and 2) each dataset can contain personalized or dataset-specific information, constituting distinct factors that are not coupled with other datasets. In this work, we introduce a personalized CTD framework tackling these challenges. A flexible model is proposed where each dataset is represented as the sum of two components, one related to a common tensor through a multilinear measurement model, and another specific to each dataset. Both the common and distinct components are assumed to admit a polyadic decomposition. This generalizes several existing CTD models. We provide conditions for specific and generic uniqueness of the decomposition that are easy to interpret. These conditions employ uni-mode uniqueness of different individual datasets and properties of the measurement model. Two algorithms are proposed to compute the common and distinct components: a semi-algebraic one and a coordinate-descent optimization method. Experimental results illustrate the advantage of the proposed framework compared with the state of the art approaches.
{"title":"Personalized Coupled Tensor Decomposition for Multimodal Data Fusion: Uniqueness and Algorithms","authors":"Ricardo A. Borsoi;Konstantin Usevich;David Brie;Tülay Adali","doi":"10.1109/TSP.2024.3510680","DOIUrl":"10.1109/TSP.2024.3510680","url":null,"abstract":"Coupled tensor decompositions (CTDs) perform data fusion by linking factors from different datasets. Although many CTDs have been already proposed, current works do not address important challenges of data fusion, where: 1) the datasets are often heterogeneous, constituting different “views” of a given phenomena (multimodality); and 2) each dataset can contain \u0000<italic>personalized</i>\u0000 or dataset-specific information, constituting distinct factors that are not coupled with other datasets. In this work, we introduce a personalized CTD framework tackling these challenges. A flexible model is proposed where each dataset is represented as the sum of two components, one related to a common tensor through a multilinear measurement model, and another specific to each dataset. Both the common and distinct components are assumed to admit a polyadic decomposition. This generalizes several existing CTD models. We provide conditions for specific and generic uniqueness of the decomposition that are easy to interpret. These conditions employ \u0000<italic>uni-mode</i>\u0000 uniqueness of different individual datasets and properties of the measurement model. Two algorithms are proposed to compute the common and distinct components: a semi-algebraic one and a coordinate-descent optimization method. Experimental results illustrate the advantage of the proposed framework compared with the state of the art approaches.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"113-129"},"PeriodicalIF":4.6,"publicationDate":"2024-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142776540","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-03DOI: 10.1109/TSP.2024.3510755
Qiyi Yao;Kai Zeng;Weiming Zhang;Kejiang Chen
Steganography is the art of covert communication that pursues the secrecy of concealment. In adaptive steganography, the most commonly used framework of steganography, the sender embeds a “secret message” signal within another “cover” signal with respect to a certain adaptive distortion function that measures the distortion incurred, contributing to the composite “stego” signal that resembles the cover, and the receiver extracts the “secret message” signal from the stego. When the communication channel between the sender and the receiver is noisy, robust steganography is needed, in which robust adaptive steganographic coding plays a central role. The existing robust adaptive steganographic coding methods can only provide very limited robustness, and they fail when the communication channel is bad. To ensure the success of covert communication, we propose a reliable robust adaptive steganographic coding scheme based on nested polar codes that possesses the highest robustness among the existing algorithms while the security performance is also maintained. Theoretically, we show that for the most important binary embedding , in the special case where the communication channel is a Binary Symmetric Channel (BSC), the proposed scheme is optimal under the constant distortion profile as the cover length $N$ tends to infinity through powers of two when the design embedding rate is large enough. Experimentally, our method is capable of making sure the perfect extraction of the secret message in situations where the embedding rate is large or the communication channel is bad, while the existing algorithms are not applicable in these scenarios.
{"title":"Reliable Robust Adaptive Steganographic Coding Based on Nested Polar Codes","authors":"Qiyi Yao;Kai Zeng;Weiming Zhang;Kejiang Chen","doi":"10.1109/TSP.2024.3510755","DOIUrl":"10.1109/TSP.2024.3510755","url":null,"abstract":"Steganography is the art of covert communication that pursues the secrecy of concealment. In adaptive steganography, the most commonly used framework of steganography, the sender embeds a “secret message” signal within another “cover” signal with respect to a certain adaptive distortion function that measures the distortion incurred, contributing to the composite “stego” signal that resembles the cover, and the receiver extracts the “secret message” signal from the stego. When the communication channel between the sender and the receiver is noisy, robust steganography is needed, in which robust adaptive steganographic coding plays a central role. The existing robust adaptive steganographic coding methods can only provide very limited robustness, and they fail when the communication channel is bad. To ensure the success of covert communication, we propose a reliable robust adaptive steganographic coding scheme based on nested polar codes that possesses the highest robustness among the existing algorithms while the security performance is also maintained. Theoretically, we show that for the most important \u0000<italic>binary embedding</i>\u0000, in the special case where the communication channel is a Binary Symmetric Channel (BSC), the proposed scheme is optimal under the constant distortion profile as the cover length \u0000<inline-formula><tex-math>$N$</tex-math></inline-formula>\u0000 tends to infinity through powers of two when the \u0000<italic>design embedding rate</i>\u0000 is large enough. Experimentally, our method is capable of making sure the perfect extraction of the secret message in situations where the embedding rate is large or the communication channel is bad, while the existing algorithms are not applicable in these scenarios.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"12-25"},"PeriodicalIF":4.6,"publicationDate":"2024-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142776569","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-03DOI: 10.1109/TSP.2024.3510434
Mohammad Bagher Iraji;Mohammad Eini;Arash Amini
In this paper, we generalize the concepts of kernels, weak stationarity and white noise from undirected to directed graphs (digraphs) based on the Jordan decomposition of the shift operator. We characterize two types of kernels (type-I and type-II) and their corresponding localization operators for digraphs. We analytically study the interplay of these types of kernels with the concept of stationarity, specially the filtering properties. We also generalize graph Wiener filters and the related optimization framework to digraphs. For the special case of Gaussian processes, we show that the Wiener filtering again coincides with the MAP estimator. We further investigate the linear minimum mean-squared error (LMMSE) estimator for the non-Gaussian cases; the corresponding optimization problem simplifies to a Lyapunov matrix equation. We propose an algorithm to solve the Wiener optimization using proximal splitting methods. Finally, we provide simulation results to verify the provided theory.
{"title":"WSS Processes and Wiener Filters on Digraphs","authors":"Mohammad Bagher Iraji;Mohammad Eini;Arash Amini","doi":"10.1109/TSP.2024.3510434","DOIUrl":"10.1109/TSP.2024.3510434","url":null,"abstract":"In this paper, we generalize the concepts of kernels, weak stationarity and white noise from undirected to directed graphs (digraphs) based on the Jordan decomposition of the shift operator. We characterize two types of kernels (type-I and type-II) and their corresponding localization operators for digraphs. We analytically study the interplay of these types of kernels with the concept of stationarity, specially the filtering properties. We also generalize graph Wiener filters and the related optimization framework to digraphs. For the special case of Gaussian processes, we show that the Wiener filtering again coincides with the MAP estimator. We further investigate the linear minimum mean-squared error (LMMSE) estimator for the non-Gaussian cases; the corresponding optimization problem simplifies to a Lyapunov matrix equation. We propose an algorithm to solve the Wiener optimization using proximal splitting methods. Finally, we provide simulation results to verify the provided theory.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"1-11"},"PeriodicalIF":4.6,"publicationDate":"2024-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142776539","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-02DOI: 10.1109/TSP.2024.3509413
Hao Jiang;Xiaojun Yuan;Qinghua Guo
Generalized bilinear factorization (GBF), in which two matrices are recovered from noisy and typically compressed measurements of their product, arises in various applications such as blind channel-and-signal estimation, image completion, and compressed video foreground and background separation. In this paper, we formulate the GBF problem by unifying several existing bilinear inverse problems, and establish a novel hybrid vector message passing (HVMP) algorithm for GBF. The GBF-HVMP algorithm integrates expectation propagation (EP) and variational message passing (VMP) via variational free energy minimization, and exchanges matrix-variable messages in closed form. GBF-HVMP is advantageous over its counterparts in several aspects. For example, a matrix-variable message can characterize the correlations between the elements of the matrix, which is not possible in scalar-variable message passing; the hybrid of EP and VMP yields closed-form Gaussian messages associated with the bilinear constraints inherent in the GBF problem. We show that damping is unnecessary for GBF-HVMP to ensure convergence. We also show that GBF-HVMP performs close to the replica bound, and significantly outperforms state-of-the-art approaches in terms of both normalized mean squared error (NMSE) performance and computational complexity.
{"title":"Generalized Bilinear Factorization via Hybrid Vector Message Passing","authors":"Hao Jiang;Xiaojun Yuan;Qinghua Guo","doi":"10.1109/TSP.2024.3509413","DOIUrl":"10.1109/TSP.2024.3509413","url":null,"abstract":"Generalized bilinear factorization (GBF), in which two matrices are recovered from noisy and typically compressed measurements of their product, arises in various applications such as blind channel-and-signal estimation, image completion, and compressed video foreground and background separation. In this paper, we formulate the GBF problem by unifying several existing bilinear inverse problems, and establish a novel hybrid vector message passing (HVMP) algorithm for GBF. The GBF-HVMP algorithm integrates expectation propagation (EP) and variational message passing (VMP) via variational free energy minimization, and exchanges matrix-variable messages in closed form. GBF-HVMP is advantageous over its counterparts in several aspects. For example, a matrix-variable message can characterize the correlations between the elements of the matrix, which is not possible in scalar-variable message passing; the hybrid of EP and VMP yields closed-form Gaussian messages associated with the bilinear constraints inherent in the GBF problem. We show that damping is unnecessary for GBF-HVMP to ensure convergence. We also show that GBF-HVMP performs close to the replica bound, and significantly outperforms state-of-the-art approaches in terms of both normalized mean squared error (NMSE) performance and computational complexity.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"72 ","pages":"5675-5690"},"PeriodicalIF":4.6,"publicationDate":"2024-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142760268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-02DOI: 10.1109/tsp.2024.3510623
Yue Hu, Songkang Huang, Lei Zhao, Ming Jiang
{"title":"Narrowband Interference Cancellation for OFDM Based on Deep Learning and Compressed Sensing","authors":"Yue Hu, Songkang Huang, Lei Zhao, Ming Jiang","doi":"10.1109/tsp.2024.3510623","DOIUrl":"https://doi.org/10.1109/tsp.2024.3510623","url":null,"abstract":"","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"26 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2024-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142760269","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-29DOI: 10.1109/TSP.2024.3508468
Joaquin Ruiz;Hau-Tieng Wu;Marcelo A. Colominas
Dealing with time series with missing values, including those afflicted by low quality or over-saturation, presents a significant signal processing challenge. The task of recovering these missing values, known as imputation, has led to the development of several algorithms. However, we have observed that the efficacy of these algorithms tends to diminish when the time series exhibits non-stationary oscillatory behavior. In this paper, we introduce a novel algorithm, coined Harmonic Level Interpolation ( HaLI ), which enhances the performance of existing imputation algorithms for oscillatory time series. After running any chosen imputation algorithm, HaLI leverages the harmonic decomposition based on the adaptive non-harmonic model of the initial imputation to improve the imputation accuracy for oscillatory time series. Experimental assessments conducted on synthetic and real signals consistently highlight that HaLI enhances the performance of existing imputation algorithms. The algorithm is made publicly available as a readily employable Matlab code for other researchers to use.
{"title":"Enhancing Missing Data Imputation of Non-Stationary Oscillatory Signals With Harmonic Decomposition","authors":"Joaquin Ruiz;Hau-Tieng Wu;Marcelo A. Colominas","doi":"10.1109/TSP.2024.3508468","DOIUrl":"10.1109/TSP.2024.3508468","url":null,"abstract":"Dealing with time series with missing values, including those afflicted by low quality or over-saturation, presents a significant signal processing challenge. The task of recovering these missing values, known as imputation, has led to the development of several algorithms. However, we have observed that the efficacy of these algorithms tends to diminish when the time series exhibits non-stationary oscillatory behavior. In this paper, we introduce a novel algorithm, coined Harmonic Level Interpolation (\u0000<monospace>HaLI</monospace>\u0000), which enhances the performance of existing imputation algorithms for oscillatory time series. After running any chosen imputation algorithm, \u0000<monospace>HaLI</monospace>\u0000 leverages the \u0000<italic>harmonic decomposition</i>\u0000 based on the \u0000<italic>adaptive non-harmonic model</i>\u0000 of the initial imputation to improve the imputation accuracy for oscillatory time series. Experimental assessments conducted on synthetic and real signals consistently highlight that \u0000<monospace>HaLI</monospace>\u0000 enhances the performance of existing imputation algorithms. The algorithm is made publicly available as a readily employable Matlab code for other researchers to use.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"72 ","pages":"5581-5592"},"PeriodicalIF":4.6,"publicationDate":"2024-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142753504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-28DOI: 10.1109/TSP.2024.3507715
Mert Kayaalp;Yunus İnan;Visa Koivunen;Ali H. Sayed
In this paper, we consider a setting where heterogeneous agents with connectivity are performing inference using unlabeled streaming data. Observed data are only partially informative about the target variable of interest. In order to overcome the uncertainty, agents cooperate with each other by exchanging their local inferences with and through a fusion center. To evaluate how each agent influences the overall decision, we adopt a causal framework in order to distinguish the actual influence of agents from mere correlations within the decision-making process. Various scenarios reflecting different agent participation patterns and fusion center policies are investigated. We derive expressions to quantify the causal impact of each agent on the joint decision, which could be beneficial for anticipating and addressing atypical scenarios, such as adversarial attacks or system malfunctions. We validate our theoretical results with numerical simulations and a real-world application of multi-camera crowd counting.
{"title":"Causal Influence in Federated Edge Inference","authors":"Mert Kayaalp;Yunus İnan;Visa Koivunen;Ali H. Sayed","doi":"10.1109/TSP.2024.3507715","DOIUrl":"10.1109/TSP.2024.3507715","url":null,"abstract":"In this paper, we consider a setting where heterogeneous agents with connectivity are performing inference using unlabeled streaming data. Observed data are only partially informative about the target variable of interest. In order to overcome the uncertainty, agents cooperate with each other by exchanging their local inferences with and through a fusion center. To evaluate how each agent influences the overall decision, we adopt a causal framework in order to distinguish the actual influence of agents from mere correlations within the decision-making process. Various scenarios reflecting different agent participation patterns and fusion center policies are investigated. We derive expressions to quantify the causal impact of each agent on the joint decision, which could be beneficial for anticipating and addressing atypical scenarios, such as adversarial attacks or system malfunctions. We validate our theoretical results with numerical simulations and a real-world application of multi-camera crowd counting.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"72 ","pages":"5604-5615"},"PeriodicalIF":4.6,"publicationDate":"2024-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142752853","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-28DOI: 10.1109/TSP.2024.3507819
Petre Stoica;Prabhu Babu
In a recent paper we have proposed an approach for estimating the covariance matrix from a multivariate data set ${mathbf{y}(t)}$ that may contain outliers. If $mathbf{y}(t)$ is flagged as outlying by this approach, then the entire vector $mathbf{y}(t)$ is considered to contain no useful information and it is discarded. However, in some applications the data contains cell outliers, that is to say, not all elements of $mathbf{y}(t)$ are outlying but only some of them. One then wants to eliminate only the cell outliers from the data, rather than the entire vector $mathbf{y}(t)$ . In this paper, we propose a penalized maximum likelihood approach to outlier detection and covariance matrix estimation from data with cell outliers. Specifically we estimate the positions of the outliers in the data set, for a given estimate of the covariance matrix, by maximizing the penalized likelihood of the data with the penalty being derived from a property of the likelihood ratio and the false discovery rate (FDR) principle. We alternate this step with a majorization-minimization (MM) technique that estimates the covariance matrix for given outlier positions. The MM is more flexible than the expectation maximization (EM) algorithm commonly used for estimating the covariance matrix from data with missing cells, as the former can be utilized in cases in which the latter is not usable. The closest competitor of our approach is the cellMCD (minimum covariance determinant) method, compared with which the proposed approach has a number of advantages described in the introduction and the numerical study section.
{"title":"Penalized Likelihood Approach to Covariance Matrix Estimation From Data With Cell Outliers","authors":"Petre Stoica;Prabhu Babu","doi":"10.1109/TSP.2024.3507819","DOIUrl":"10.1109/TSP.2024.3507819","url":null,"abstract":"In a recent paper we have proposed an approach for estimating the covariance matrix from a multivariate data set \u0000<inline-formula><tex-math>${mathbf{y}(t)}$</tex-math></inline-formula>\u0000 that may contain outliers. If \u0000<inline-formula><tex-math>$mathbf{y}(t)$</tex-math></inline-formula>\u0000 is flagged as outlying by this approach, then the entire vector \u0000<inline-formula><tex-math>$mathbf{y}(t)$</tex-math></inline-formula>\u0000 is considered to contain no useful information and it is discarded. However, in some applications the data contains cell outliers, that is to say, not all elements of \u0000<inline-formula><tex-math>$mathbf{y}(t)$</tex-math></inline-formula>\u0000 are outlying but only some of them. One then wants to eliminate only the cell outliers from the data, rather than the entire vector \u0000<inline-formula><tex-math>$mathbf{y}(t)$</tex-math></inline-formula>\u0000. In this paper, we propose a penalized maximum likelihood approach to outlier detection and covariance matrix estimation from data with cell outliers. Specifically we estimate the positions of the outliers in the data set, for a given estimate of the covariance matrix, by maximizing the penalized likelihood of the data with the penalty being derived from a property of the likelihood ratio and the false discovery rate (FDR) principle. We alternate this step with a majorization-minimization (MM) technique that estimates the covariance matrix for given outlier positions. The MM is more flexible than the expectation maximization (EM) algorithm commonly used for estimating the covariance matrix from data with missing cells, as the former can be utilized in cases in which the latter is not usable. The closest competitor of our approach is the cellMCD (minimum covariance determinant) method, compared with which the proposed approach has a number of advantages described in the introduction and the numerical study section.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"72 ","pages":"5616-5627"},"PeriodicalIF":4.6,"publicationDate":"2024-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142752854","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-28DOI: 10.1109/TSP.2024.3507572
Georgios Apostolakis;Aggelos Bletsas
Distributed execution of algorithms over various terminals is a topic that regains increasing popularity; when tolerance to failures is also required, asynchronous operation is brought to the light, while probabilistic asynchronous operation can model the probability of failure for each terminal. This work focuses on the probabilistic asynchronous affine update model, applicable in a wide range of inference algorithms, possibly executed over distributed terminals. The existing literature focuses on the asymptotic properties of the expected mean. Instead, this work offers the asymptotic analysis for the arithmetic mean, utilized for discovering fixed points, as it is the only quantity that can be practically offered experimentally. It is shown that the asymptotic behavior of the arithmetic mean is different than the expected mean's and a sufficient condition is provided for convergence of the arithmetic mean to a fixed point. The lack of necessity for this condition is explained and the subcases, where the arithmetic mean converges, diverges or has an unpredictable behavior, are distinguished. Additionally, cases where the individual iterations never converge (e.g., oscillate infinitely) but their arithmetic mean does and offers fixed point, are also highlighted. This is another concrete example of the arithmetic mean utility. Applications of the affine model are also briefly discussed. Finally, simulations corroborate theoretical findings for various affine model setups.
{"title":"Arithmetic Vs. Expected Mean of Probabilistic Asynchronous Affine Inference","authors":"Georgios Apostolakis;Aggelos Bletsas","doi":"10.1109/TSP.2024.3507572","DOIUrl":"10.1109/TSP.2024.3507572","url":null,"abstract":"Distributed execution of algorithms over various terminals is a topic that regains increasing popularity; when tolerance to failures is also required, asynchronous operation is brought to the light, while probabilistic asynchronous operation can model the probability of failure for each terminal. This work focuses on the probabilistic asynchronous affine update model, applicable in a wide range of inference algorithms, possibly executed over distributed terminals. The existing literature focuses on the asymptotic properties of the expected mean. Instead, this work offers the asymptotic analysis for the arithmetic mean, utilized for discovering fixed points, as it is the only quantity that can be practically offered experimentally. It is shown that the asymptotic behavior of the arithmetic mean is different than the expected mean's and a sufficient condition is provided for convergence of the arithmetic mean to a fixed point. The lack of necessity for this condition is explained and the subcases, where the arithmetic mean converges, diverges or has an unpredictable behavior, are distinguished. Additionally, cases where the individual iterations never converge (e.g., oscillate infinitely) but their arithmetic mean does and offers fixed point, are also highlighted. This is another concrete example of the arithmetic mean utility. Applications of the affine model are also briefly discussed. Finally, simulations corroborate theoretical findings for various affine model setups.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"72 ","pages":"5593-5603"},"PeriodicalIF":4.6,"publicationDate":"2024-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142752852","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-27DOI: 10.1109/TSP.2024.3505938
You Xu;Guanghua Liu;Xiaotong Lu;Chao Xie;Lixia Xiao;Tao Jiang
Non-line-of-sight (NLOS) mmWave imaging technology reconstructs the contour features of hidden targets by analyzing the indirect reflected signals of the relay surface, which has been a hot topic in disaster reserve and autonomous driving. However, due to the differences in the reflecting characteristics of multiangle relay surfaces, traditional multipath utilization methods inevitably suffer from disturbance, and obtaining high-quality images remains a challenging task. In this paper, we propose a double sparse structure enhanced mmWave NLOS imaging framework. First, we establish an automotive-squint synthetic aperture radar (AS-SAR) model under multiangle relay surface and analyze the multiangle image characteristics. Subsequently, we introduce a double sparse structure to transform the image reconstruction problem into a hybrid convex regularization problem, and theoretically derive the minimum lower bounds of sample complexity and estimation error. Then, based on the fast iterative threshold shrinkage framework, we propose a time-domain double sparse thresholding algorithm (TD-DSTA), in which the double sparse operator is optimized by alternating direction multiplication. In addition, we propose a two-dimensional frequency domain method based on the approximate-operator to reduce the computational complexity. Finally, we evaluate the performance of the proposed method through quantitative and qualitative analysis in the NLOS multiangle relay surfaces scenario. Simulation and real experimental results verify the superiority of the proposed method in NLOS image reconstruction.
{"title":"Double Sparse Structure-Enhanced mmWave NLOS Imaging Under Multiangle Relay Surface","authors":"You Xu;Guanghua Liu;Xiaotong Lu;Chao Xie;Lixia Xiao;Tao Jiang","doi":"10.1109/TSP.2024.3505938","DOIUrl":"10.1109/TSP.2024.3505938","url":null,"abstract":"Non-line-of-sight (NLOS) mmWave imaging technology reconstructs the contour features of hidden targets by analyzing the indirect reflected signals of the relay surface, which has been a hot topic in disaster reserve and autonomous driving. However, due to the differences in the reflecting characteristics of multiangle relay surfaces, traditional multipath utilization methods inevitably suffer from disturbance, and obtaining high-quality images remains a challenging task. In this paper, we propose a double sparse structure enhanced mmWave NLOS imaging framework. First, we establish an automotive-squint synthetic aperture radar (AS-SAR) model under multiangle relay surface and analyze the multiangle image characteristics. Subsequently, we introduce a double sparse structure to transform the image reconstruction problem into a hybrid convex regularization problem, and theoretically derive the minimum lower bounds of sample complexity and estimation error. Then, based on the fast iterative threshold shrinkage framework, we propose a time-domain double sparse thresholding algorithm (TD-DSTA), in which the double sparse operator is optimized by alternating direction multiplication. In addition, we propose a two-dimensional frequency domain method based on the approximate-operator to reduce the computational complexity. Finally, we evaluate the performance of the proposed method through quantitative and qualitative analysis in the NLOS multiangle relay surfaces scenario. Simulation and real experimental results verify the superiority of the proposed method in NLOS image reconstruction.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"72 ","pages":"5628-5643"},"PeriodicalIF":4.6,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142753466","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：