Pub Date : 2025-11-26DOI: 10.1109/TSP.2025.3637317
Niclas Führling;Giuseppe Thadeu Freitas de Abreu;David González G.;Osvaldo Gonsa
We consider a robust and self-reliant (or “egoistic”) variation of the rigid body localization (RBL) problem, in which a primary rigid body seeks to estimate the pose ($i.e.$, location and orientation) of another rigid body (or “target”), relative to its own, without the assistance of external infrastructure, without prior knowledge of the shape of the target, and taking into account the possibility that the available observations are incomplete. Three complementary contributions are then offered for such a scenario. The first is a method to estimate the translation vector between the center points of both rigid bodies, which unlike existing techniques does not require that both objects have the same shape or even the same number of landmark points. This technique is shown to significantly outperform the state-of-the-art (SotA) under complete information, but to be sensitive to data erasures, even when enhanced by matrix completion methods. The second contribution, designed to offer improved performance in the presence of incomplete information, offers a robust alternative to the latter, at the expense of a slight relative loss under complete information. Finally, the third contribution is a scheme for the estimation of the rotation matrix describing the relative orientation of the target rigid body with respect to the primary. Comparisons of the proposed schemes and SotA techniques demonstrate the advantage of the contributed methods in terms of root mean square error (RMSE) performance under fully complete information and incomplete conditions.
{"title":"Robust Egoistic Rigid Body Localization","authors":"Niclas Führling;Giuseppe Thadeu Freitas de Abreu;David González G.;Osvaldo Gonsa","doi":"10.1109/TSP.2025.3637317","DOIUrl":"10.1109/TSP.2025.3637317","url":null,"abstract":"We consider a robust and self-reliant (or “egoistic”) variation of the rigid body localization (RBL) problem, in which a primary rigid body seeks to estimate the pose (<inline-formula><tex-math>$i.e.$</tex-math></inline-formula>, location and orientation) of another rigid body (or “target”), relative to its own, without the assistance of external infrastructure, without prior knowledge of the shape of the target, and taking into account the possibility that the available observations are incomplete. Three complementary contributions are then offered for such a scenario. The first is a method to estimate the translation vector between the center points of both rigid bodies, which unlike existing techniques does not require that both objects have the same shape or even the same number of landmark points. This technique is shown to significantly outperform the state-of-the-art (SotA) under complete information, but to be sensitive to data erasures, even when enhanced by matrix completion methods. The second contribution, designed to offer improved performance in the presence of incomplete information, offers a robust alternative to the latter, at the expense of a slight relative loss under complete information. Finally, the third contribution is a scheme for the estimation of the rotation matrix describing the relative orientation of the target rigid body with respect to the primary. Comparisons of the proposed schemes and SotA techniques demonstrate the advantage of the contributed methods in terms of root mean square error (RMSE) performance under fully complete information and incomplete conditions.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"5076-5089"},"PeriodicalIF":5.8,"publicationDate":"2025-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145609347","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 : 2025-11-26DOI: 10.1109/TSP.2025.3637278
Hongqi Min;Xinrui Li;Ruoguang Li;Yong Zeng
For the sixth generation (6G) wireless networks, achieving high-performance integrated localization and communication (ILAC) is critical to unlock the full potential of wireless networks. To simultaneously enhance wireless localization and communication performance cost-effectively, this paper proposes sparse multiple-input multiple-output (MIMO) based ILAC with nested and co-prime sparse arrays deployed at the base station (BS). Sparse MIMO relaxes the traditional half-wavelength antenna spacing constraint to enlarge the antenna aperture, thus enhancing localization degrees of freedom (DoFs) and providing finer spatial resolution. However, it also leads to undesired grating lobes, which may cause severe inter-user interference (IUI) for communication and angular ambiguity for localization. While the latter issue can be effectively addressed by the virtual array technology, by forming sum or difference co-arrays via signal (conjugate) correlation among array elements, it is unclear whether the similar virtual array technology also benefits wireless communications for ILAC systems. In this paper, we first reveal that the answer to the above question is negative, by showing that forming virtual arrays for wireless communication will cause destruction of phase information, degradation of signal-to-noise ratio and aggravation of multi-user interference. Therefore, we propose the so-called hybrid processing framework for sparse MIMO based ILAC, i.e., physical array based communication while virtual array based localization. To this end, we characterize the beam pattern of sparse arrays by three metrics, i.e., main lobe beam width, peak-to-local-minimum ratio, and side lobe height, demonstrating that despite of the undesired grating lobes, sparse arrays can also bring benefits to communications, thanks to its narrower main lobe beam width than the conventional compact arrays. Extensive simulation results are presented to demonstrate the performance gains of sparse MIMO based ILAC over that based on the conventional compact MIMO.
{"title":"Integrated Localization and Communication With Sparse MIMO: Will Virtual Array Technology Also Benefit Wireless Communication?","authors":"Hongqi Min;Xinrui Li;Ruoguang Li;Yong Zeng","doi":"10.1109/TSP.2025.3637278","DOIUrl":"10.1109/TSP.2025.3637278","url":null,"abstract":"For the sixth generation (6G) wireless networks, achieving high-performance integrated localization and communication (ILAC) is critical to unlock the full potential of wireless networks. To simultaneously enhance wireless localization and communication performance cost-effectively, this paper proposes sparse multiple-input multiple-output (MIMO) based ILAC with nested and co-prime sparse arrays deployed at the base station (BS). Sparse MIMO relaxes the traditional half-wavelength antenna spacing constraint to enlarge the antenna aperture, thus enhancing localization degrees of freedom (DoFs) and providing finer spatial resolution. However, it also leads to undesired grating lobes, which may cause severe inter-user interference (IUI) for communication and angular ambiguity for localization. While the latter issue can be effectively addressed by the virtual array technology, by forming sum or difference co-arrays via signal (conjugate) correlation among array elements, it is unclear whether the similar virtual array technology also benefits wireless communications for ILAC systems. In this paper, we first reveal that the answer to the above question is negative, by showing that forming virtual arrays for wireless communication will cause destruction of phase information, degradation of signal-to-noise ratio and aggravation of multi-user interference. Therefore, we propose the so-called hybrid processing framework for sparse MIMO based ILAC, i.e., physical array based communication while virtual array based localization. To this end, we characterize the beam pattern of sparse arrays by three metrics, i.e., main lobe beam width, peak-to-local-minimum ratio, and side lobe height, demonstrating that despite of the undesired grating lobes, sparse arrays can also bring benefits to communications, thanks to its narrower main lobe beam width than the conventional compact arrays. Extensive simulation results are presented to demonstrate the performance gains of sparse MIMO based ILAC over that based on the conventional compact MIMO.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"5090-5105"},"PeriodicalIF":5.8,"publicationDate":"2025-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145609421","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 : 2025-11-24DOI: 10.1109/tsp.2025.3636071
Sergio Rozada, Hoi-To Wai, Antonio G. Marques
{"title":"Multilinear Tensor Low-Rank Approximation for Policy-Gradient Methods in Reinforcement Learning","authors":"Sergio Rozada, Hoi-To Wai, Antonio G. Marques","doi":"10.1109/tsp.2025.3636071","DOIUrl":"https://doi.org/10.1109/tsp.2025.3636071","url":null,"abstract":"","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"14 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145593242","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 : 2025-11-21DOI: 10.1109/TSP.2025.3632844
Hao Shu;Jicheng Li;Yu Jin;Hailin Wang
In recent years, the prediction of multidimensional time series data has become increasingly important due to its wide-ranging applications. Tensor-based prediction methods have gained attention for their ability to preserve the inherent structure of such data. However, existing approaches, such as tensor autoregression and tensor decomposition, often have consistently failed to provide clear assertions regarding the number of samples that can be exactly predicted. While matrix-based methods using nuclear norms address this limitation, their reliance on matrices limits accuracy and increases computational costs when handling multidimensional data. To overcome these challenges, we reformulate multidimensional time series prediction as a deterministic tensor completion problem and propose a novel theoretical framework. Specifically, we develop a deterministic tensor completion theory and introduce the Temporal Convolutional Tensor Nuclear Norm (TCTNN) model. By convolving the multidimensional time series along the temporal dimension and applying the tensor nuclear norm, our approach identifies the maximum forecast horizon for exact predictions. Additionally, TCTNN achieves superior performance in prediction accuracy and computational efficiency compared to existing methods across diverse real-world datasets, including climate temperature, network flow, and traffic ride data. Our implementation is publicly available at https://github.com/HaoShu2000/TCTNN.
{"title":"Guaranteed Multidimensional Time Series Prediction via Deterministic Tensor Completion Theory","authors":"Hao Shu;Jicheng Li;Yu Jin;Hailin Wang","doi":"10.1109/TSP.2025.3632844","DOIUrl":"10.1109/TSP.2025.3632844","url":null,"abstract":"In recent years, the prediction of multidimensional time series data has become increasingly important due to its wide-ranging applications. Tensor-based prediction methods have gained attention for their ability to preserve the inherent structure of such data. However, existing approaches, such as tensor autoregression and tensor decomposition, often have consistently failed to provide clear assertions regarding the number of samples that can be exactly predicted. While matrix-based methods using nuclear norms address this limitation, their reliance on matrices limits accuracy and increases computational costs when handling multidimensional data. To overcome these challenges, we reformulate multidimensional time series prediction as a deterministic tensor completion problem and propose a novel theoretical framework. Specifically, we develop a deterministic tensor completion theory and introduce the <italic>Temporal Convolutional Tensor Nuclear Norm</i> (TCTNN) model. By convolving the multidimensional time series along the temporal dimension and applying the tensor nuclear norm, our approach identifies the maximum forecast horizon for exact predictions. Additionally, TCTNN achieves superior performance in prediction accuracy and computational efficiency compared to existing methods across diverse real-world datasets, including climate temperature, network flow, and traffic ride data. Our implementation is publicly available at <uri>https://github.com/HaoShu2000/TCTNN</uri>.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"4638-4653"},"PeriodicalIF":5.8,"publicationDate":"2025-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145567944","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 : 2025-11-20DOI: 10.1109/tsp.2025.3634427
Luan Portella, F. M. Bayer, R. J. Cintra
{"title":"Multiplierless DFT Approximation Based on the Prime Factor Algorithm","authors":"Luan Portella, F. M. Bayer, R. J. Cintra","doi":"10.1109/tsp.2025.3634427","DOIUrl":"https://doi.org/10.1109/tsp.2025.3634427","url":null,"abstract":"","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"6 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145559337","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 : 2025-11-20DOI: 10.1109/TSP.2025.3626640
Kai Li;Wenqiang Pu;Zhi-Quan Luo
In current 5G massive multiple-input multiple-output (MIMO) cellular networks, the performance of beamforming hinges critically on the accuracy of downlink channel state information (CSI), particularly in frequency division duplexing (FDD) systems. The absence of channel reciprocity in FDD systems introduces notable challenges, resulting in substantial communication overhead when transmitting downlink CSI directly. To tackle this obstacle, a limited feedback strategy is adopted to compress the downlink CSI into a manageable number of bits. Nevertheless, this compression complicates the accurate retrieval of downlink CSI, reducing the beamforming performance. This paper thoroughly examines the limited feedback mechanism, scrutinizing its structural design and the values it generates. Drawing upon these insights, we introduce an online algorithm that alternates between exploration and estimation to enhance beamforming for single-stream and multi-stream configurations. Our proposed method significantly elevates beamforming vector quality by adeptly decoding downlink CSI from the limited feedback. Additionally, it leverages existing information to refine the feedback process, leading to the generation of more precise CSI. As numerical results demonstrate, the complementary nature of estimation and exploration leads to outstanding performance in optimal beamforming acquisition.
{"title":"Efficient Beamforming Refinement for Limited Feedback FDD Massive MIMO: An Online Alternating Exploration-Estimation Approach","authors":"Kai Li;Wenqiang Pu;Zhi-Quan Luo","doi":"10.1109/TSP.2025.3626640","DOIUrl":"10.1109/TSP.2025.3626640","url":null,"abstract":"In current 5G massive multiple-input multiple-output (MIMO) cellular networks, the performance of beamforming hinges critically on the accuracy of downlink channel state information (CSI), particularly in frequency division duplexing (FDD) systems. The absence of channel reciprocity in FDD systems introduces notable challenges, resulting in substantial communication overhead when transmitting downlink CSI directly. To tackle this obstacle, a limited feedback strategy is adopted to compress the downlink CSI into a manageable number of bits. Nevertheless, this compression complicates the accurate retrieval of downlink CSI, reducing the beamforming performance. This paper thoroughly examines the limited feedback mechanism, scrutinizing its structural design and the values it generates. Drawing upon these insights, we introduce an online algorithm that alternates between exploration and estimation to enhance beamforming for single-stream and multi-stream configurations. Our proposed method significantly elevates beamforming vector quality by adeptly decoding downlink CSI from the limited feedback. Additionally, it leverages existing information to refine the feedback process, leading to the generation of more precise CSI. As numerical results demonstrate, the complementary nature of estimation and exploration leads to outstanding performance in optimal beamforming acquisition.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"4969-4985"},"PeriodicalIF":5.8,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145559338","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}
This paper investigates the signal activity detection problem for spread spectrum signals in low probability of detection communication systems under Gaussian disturbance with unknown covariance matrix. We assume that the receiver is equipped with an adaptive antenna array with general array configuration, comprising of a primary array with high–gain antennas and a reference array with low–gain antennas. It has been shown that this reference array structure can benefit interference suppression. Since it is difficult to derive the uniformly most powerful detector for the detection problem, we resort to using the Wald test and generalized likelihood ratio test (GLRT) schemes to design the detectors. Analytical performance of the proposed Wald test and GLRT is derived, indicating that the proposed two detectors bear constant false alarm rate property. Finally, numerical simulations are carried out, which verify the correctness of the theoretical analysis. Besides, it is revealed that the use of reference channel can facilitate to enhance the detection performance, but an increasing in the antenna number does not necessarily lead to an improvement in detection performance.
{"title":"Adaptive Detection of Spread Spectrum Signals With General Array Configuration","authors":"Yutong Feng;Akihito Taya;Yuuki Nishiyama;Jun Liu;Kaoru Sezaki","doi":"10.1109/TSP.2025.3633721","DOIUrl":"10.1109/TSP.2025.3633721","url":null,"abstract":"This paper investigates the signal activity detection problem for spread spectrum signals in low probability of detection communication systems under Gaussian disturbance with unknown covariance matrix. We assume that the receiver is equipped with an adaptive antenna array with general array configuration, comprising of a primary array with high–gain antennas and a reference array with low–gain antennas. It has been shown that this reference array structure can benefit interference suppression. Since it is difficult to derive the uniformly most powerful detector for the detection problem, we resort to using the Wald test and generalized likelihood ratio test (GLRT) schemes to design the detectors. Analytical performance of the proposed Wald test and GLRT is derived, indicating that the proposed two detectors bear constant false alarm rate property. Finally, numerical simulations are carried out, which verify the correctness of the theoretical analysis. Besides, it is revealed that the use of reference channel can facilitate to enhance the detection performance, but an increasing in the antenna number does not necessarily lead to an improvement in detection performance.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"4826-4839"},"PeriodicalIF":5.8,"publicationDate":"2025-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145545613","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 : 2025-11-17DOI: 10.1109/TSP.2025.3632064
Xiaomeng Chen;Huiwen Yang;Subhrakanti Dey;Ling Shi
Distributed non-cooperative games are prevalent in emerging applications such as traffic control, vehicle charging, and smart grid management. In distributed systems without central coordinators, agents must share and retrieve information locally to seek a Nash equilibrium (NE). However, this extensive data exchange can lead to significant communication bottlenecks. To address this challenge, over-the-air computing provides a promising solution by exploiting the superposition property of wireless multiple access channels (MAC), allowing for substantial bandwidth savings. In this paper, we propose an over-the-air framework for general distributed non-cooperative games. Specifically, we introduce an algorithm based on non-coherent over-the-air computing, AirNES, to find an NE in distributed non-cooperative games. Our algorithm accounts for noisy channels and non-coherent transmission, eliminating the need for channel state information. We demonstrate that, with properly tuned decreasing consensus and gradient stepsizes, AirNES guarantees almost sure convergence to the exact NE, even in the presence of channel fading and additive noise. Additionally, we extend our analysis to scenarios with fixed stepsizes, where linear convergence can be achieved at the expense of reduced accuracy. Finally, we provide numerical simulations to demonstrate the effectiveness of the proposed protocol.
{"title":"Non-Coherent Over-the-Air Decentralized Method for Non-Cooperative Games in Multi-Agent Systems","authors":"Xiaomeng Chen;Huiwen Yang;Subhrakanti Dey;Ling Shi","doi":"10.1109/TSP.2025.3632064","DOIUrl":"10.1109/TSP.2025.3632064","url":null,"abstract":"Distributed non-cooperative games are prevalent in emerging applications such as traffic control, vehicle charging, and smart grid management. In distributed systems without central coordinators, agents must share and retrieve information locally to seek a Nash equilibrium (NE). However, this extensive data exchange can lead to significant communication bottlenecks. To address this challenge, over-the-air computing provides a promising solution by exploiting the superposition property of wireless multiple access channels (MAC), allowing for substantial bandwidth savings. In this paper, we propose an over-the-air framework for general distributed non-cooperative games. Specifically, we introduce an algorithm based on non-coherent over-the-air computing, AirNES, to find an NE in distributed non-cooperative games. Our algorithm accounts for noisy channels and non-coherent transmission, eliminating the need for channel state information. We demonstrate that, with properly tuned decreasing consensus and gradient stepsizes, AirNES guarantees almost sure convergence to the exact NE, even in the presence of channel fading and additive noise. Additionally, we extend our analysis to scenarios with fixed stepsizes, where linear convergence can be achieved at the expense of reduced accuracy. Finally, we provide numerical simulations to demonstrate the effectiveness of the proposed protocol.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"4684-4699"},"PeriodicalIF":5.8,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145535611","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}
Machine learning (ML) models are often sensitive to carefully crafted yet seemingly unnoticeable perturbations. Such adversarial examples are considered to be a property of machine learning (ML) models, often associated with their black-box operation and sensitivity to features learned from data. This work examines the adversarial sensitivity of non-learned decision rules, and particularly of iterative optimizers. Our analysis is inspired by the recent developments in deep unfolding, which cast such optimizers as ML models. We show that non-learned iterative optimizers share the sensitivity to adversarial examples of ML models, and that attacking iterative optimizers effectively alters the optimization objective surface in a manner that modifies the minima sought. We then leverage the ability to cast iteration-limited optimizers as ML models to enhance robustness via adversarial training. For a class of proximal gradient optimizers, we rigorously prove how their learning affects adversarial sensitivity. We numerically back our findings, showing the vulnerability of various optimizers, as well as the robustness induced by unfolding and adversarial training.
{"title":"Unveiling and Mitigating Adversarial Vulnerabilities in Iterative Optimizers","authors":"Elad Sofer;Tomer Shaked;Caroline Chaux;Nir Shlezinger","doi":"10.1109/TSP.2025.3633304","DOIUrl":"10.1109/TSP.2025.3633304","url":null,"abstract":"Machine learning (ML) models are often sensitive to carefully crafted yet seemingly unnoticeable perturbations. Such adversarial examples are considered to be a property of machine learning (ML) models, often associated with their black-box operation and sensitivity to features learned from data. This work examines the adversarial sensitivity of non-learned decision rules, and particularly of iterative optimizers. Our analysis is inspired by the recent developments in deep unfolding, which cast such optimizers as ML models. We show that non-learned iterative optimizers share the sensitivity to adversarial examples of ML models, and that attacking iterative optimizers effectively alters the optimization objective surface in a manner that modifies the minima sought. We then leverage the ability to cast iteration-limited optimizers as ML models to enhance robustness via adversarial training. For a class of proximal gradient optimizers, we rigorously prove how their learning affects adversarial sensitivity. We numerically back our findings, showing the vulnerability of various optimizers, as well as the robustness induced by unfolding and adversarial training.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"4669-4683"},"PeriodicalIF":5.8,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145535610","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 : 2025-11-17DOI: 10.1109/TSP.2025.3633496
Tiancheng Li;Jingyuan Wang;Guchong Li;Dengwei Gao
To solve the target tracking problem with little a-priori information about the target dynamics, our series of studies, including this paper as the third part, propose a continuous-time trajectory estimation approach (dubbed targeting track) based on the stochastic process (SP) theory and a deterministic-stochastic decomposition framework. Specifically, we decompose the learning of the trajectory SP into two sequential stages: the first fits the deterministic trend of the trajectory using a curve function of time, while the second estimates the residual stochastic component through learning either a Gaussian process (GP) or Student’s-$t$ process (StP). The former has been addressed in the companion paper and the latter is the focus of this paper. This leads to a data-driven tracking approach that produces the continuous-time trajectory with minimal prior knowledge of the target dynamics. Notably, our approach models the temporal correlations of the state sequence and of measurement noise using separate GP or StP. It does not only take advantage of the smooth trend of the target but also makes use of the long-term temporal correlation of both the data and the model fitting error. Although the GP admits an exact closed-form expression for the linear system, approximations have to be adopted for StP modeling. Simulations in four maneuvering target tracking scenarios have demonstrated its effectiveness and superiority in comparison with existing approaches.
{"title":"From Target Tracking to Targeting Track — Part III: Stochastic Process Modeling and Online Learning","authors":"Tiancheng Li;Jingyuan Wang;Guchong Li;Dengwei Gao","doi":"10.1109/TSP.2025.3633496","DOIUrl":"10.1109/TSP.2025.3633496","url":null,"abstract":"To solve the target tracking problem with little a-priori information about the target dynamics, our series of studies, including this paper as the third part, propose a continuous-time trajectory estimation approach (dubbed targeting track) based on the stochastic process (SP) theory and a deterministic-stochastic decomposition framework. Specifically, we decompose the learning of the trajectory SP into two sequential stages: the first fits the deterministic trend of the trajectory using a curve function of time, while the second estimates the residual stochastic component through learning either a Gaussian process (GP) or Student’s-<inline-formula><tex-math>$t$</tex-math></inline-formula> process (StP). The former has been addressed in the companion paper and the latter is the focus of this paper. This leads to a data-driven tracking approach that produces the continuous-time trajectory with minimal prior knowledge of the target dynamics. Notably, our approach models the temporal correlations of the state sequence and of measurement noise using separate GP or StP. It does not only take advantage of the smooth trend of the target but also makes use of the long-term temporal correlation of both the data and the model fitting error. Although the GP admits an exact closed-form expression for the linear system, approximations have to be adopted for StP modeling. Simulations in four maneuvering target tracking scenarios have demonstrated its effectiveness and superiority in comparison with existing approaches.","PeriodicalId":13330,"journal":{"name":"IEEE Transactions on Signal Processing","volume":"73 ","pages":"5336-5347"},"PeriodicalIF":5.8,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145535609","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：