One of the main challenges of federated learning (FL) is handling non-independent and identically distributed (non-IID) client data, which may occur in practice due to unbalanced datasets and use of different data sources across clients. Knowledge sharing and model personalization are key strategies for addressing this issue. Clustered federated learning is a class of FL methods that groups clients that observe similarly distributed data into clusters, such that every client is typically associated with one data distribution and participates in training a model for that distribution along their cluster peers. In this paper, we present a unified Bayesian framework for clustered FL which associates clients to clusters. Then we propose several practical algorithms to handle the, otherwise growing, data associations in a way that trades off performance and computational complexity. This work provides insights on client-cluster associations and enables client knowledge sharing in new ways. The proposed framework circumvents the need for unique client-cluster associations, which is seen to increase the performance of the resulting models in a variety of experiments.

Autonomous agents are increasingly deployed in dynamic environments where their ability to perform a given task depends on both individual and collective proficiency. While PSA has been studied for single agents, its extension to a team of agents remains underexplored. This letter addresses this gap by introducing a framework for team PSA in centralized settings. Specifically, we investigate two metrics for centralized team PSA: the MPB and the KS statistic. These metrics quantify the in-situ discrepancy between predicted and actual measurements. Then, we use the KL divergence as a reference metric. Simulations in a target tracking scenario demonstrate that both MPB and KS metrics accurately capture model mismatches, align with the KL divergence reference, and enable real-time proficiency assessment.

State-space estimation and tracking rely on accurate dynamical models to perform well. However, obtaining an accurate dynamical model for complex scenarios or adapting to changes in the system poses challenges to the estimation process. Recently, augmented physics-based models (APBMs) appear as an appealing strategy to cope with these challenges where the composition of a small and adaptive neural network with known physics-based models (PBM) is learned on the fly following an augmented state-space estimation approach. A major issue when introducing data-driven components in such a scenario is the danger of compromising the meaning (or interpretability) of estimated states. In this work, we propose a novel constrained estimation strategy that constrains the APBM dynamics close to the PBM. The novel state-space constrained approach leads to more flexible ways to impose constraints than the traditional APBM approach. Our experiments with a radar-tracking scenario demonstrate different aspects of the proposed approach and the trade-offs inherent in the imposed constraints.

Accurate localization in dense urban areas remains a significant challenge due to the limitations of Global Navigation Satellite Systems (GNSS) in environments with obstacles and reflections, such as urban canyons. While the most recent 3GPP standards offer sophisticated network-centric positioning techniques, their widespread deployment will take time and is hindered by high infrastructure costs and complexity. In this work, we present mm-NOLOC, a UE-centric localization system, designed as a practical fallback when GNSS fails to deliver high accuracy, that leverages the growing deployment of 5G mmWave infrastructure in dense urban areas. Unlike traditional approaches, mm-NOLOC operates independently of 3GPP location support and utilizes only standardized control-plane information collected solely on the UE side – Synchronization Signal Block (SSB) Indices that are mapped to 5G mmWave beam directions – to obtain robust position estimations. To address the uncertainty introduced by urban multipath, mm-NOLOC models the SSB-to-angle relationship as a discrete and multimodal distribution, based on empirical measurements in operational 5G mmWave networks, and uses a particle filter to refine position estimates by integrating probabilistic observations with UE-side motion dynamics. We validate mm-NOLOC through experiments over commercial 5G mmWave deployments, as well as trace-based simulations. Our results show that mm-NOLOC achieves a median localization error below 3 m and a 95th percentile error below 10 m, offering a practical fallback localization solution in urban canyon scenarios for 5G networks without network location support.

This article addresses near-field localization using Reconfigurable Intelligent Surfaces (RIS) in 5G systems, where Line-of-Sight (LOS) between the base station and the user is obstructed. We propose a RIS phase optimization method based on the minimization of the Cramér-Rao Lower Bound (CRLB) for position estimation. The main contributions of the article are: (1) the derivation of the CRLB to optimize the RIS configuration; and (2) the application of the mentioned framework in nearfield considering reflective RIS. The proposed method is validated with simulations, showing an accuracy improvement of RIS phase optimization with respect to the state-of-the-art methods.

This special session paper introduces the Horizon Europe NEUROPULS project, which targets the development of secure and energy-efficient RISC-V interfaced neuromorphic accelerators using augmented silicon photonics technology. Our approach aims to develop an augmented silicon photonics platform, an FPGA-powered RISC-V-connected computing platform, and a complete simulation platform to demonstrate the neuromorphic accelerator capabilities. In particular, their main advantages and limitations will be addressed concerning the underpinning technology for each platform. Then, we will discuss three targeted use cases for edge-computing applications: Global National Satellite System (GNSS) anti-jamming, autonomous driving, and anomaly detection in edge devices. Finally, we will address the reliability and security aspects of the stand-alone accelerator implementation and the project use cases.

Monitoring interferences to satellite-based navigation systems is of paramount importance in order to reliably operate critical infrastructures, navigation systems, and a variety of applications relying on satellite-based positioning. This paper investigates the use of crowd-sourced data to achieve such detection and monitoring at a central node that receives the data from an arbitrary number of agents in an area of interest. Under ideal conditions, the pathloss model is used to compute the Cramer-Rao Bound of accuracy as well as the´ corresponding maximum likelihood estimator. However, in real scenarios where obstructions and reflections are common, the signal propagation is far more complex than the pathloss model can explain. We propose to augment the pathloss model with a data-driven component, able to explain the complexities of the propagation channel. The paper shows a general methodology to jointly estimate the interference location and the parameters of the augmented model, showing superior performances in complex scenarios such as those encountered in urban environments.

A lower bound is an important tool for predicting the performance that an estimator can achieve under a particular statistical model. Bayesian bounds are a kind of such bounds which not only utilizes the observation statistics but also includes the prior model information. In reality, however, the true model generating the data is either unknown or simplified when deriving estimators, which motivates the works to derive estimation bounds under modeling mismatch situations. This paper provides a derivation of a Bayesian Cramér-Rao bound under model misspecification, by introducing important concepts such as the pseudotrue parameter in a Bayesian context which was not identified in previous works. The general result is particularized in linear and Gaussian problems, where closed-forms are available and results are used to validate the results.

Multitemporal hyperspectral unmixing (MTHU) is a fundamental tool in the analysis of hyperspectral image sequences. It reveals the dynamical evolution of the materials (endmembers) and of their proportions (abundances) in a given scene. However, adequately accounting for the spatial and temporal variability of the endmembers in MTHU is challenging, and has not been fully addressed so far in unsupervised frameworks. In this work, we propose an unsupervised MTHU algorithm based on variational recurrent neural networks. First, a stochastic model is proposed to represent both the dynamical evolution of the endmembers and their abundances, as well as the mixing process. Moreover, a new model based on a low-dimensional parametrization is used to represent spatial and temporal endmember variability, significantly reducing the amount of variables to be estimated. We propose to formulate MTHU as a Bayesian inference problem. However, the solution to this problem does not have an analytical solution due to the nonlinearity and non-Gaussianity of the model. Thus, we propose a solution based on deep variational inference, in which the posterior distribution of the estimated abundances and endmembers is represented by using a combination of recurrent neural networks and a physically motivated model. The parameters of the model are learned using stochastic backpropagation. Experimental results show that the proposed method outperforms state of the art MTHU algorithms.

The recent evolution of hyperspectral imaging technology and the proliferation of new emerging applications press for the processing of multiple temporal hyperspectral images. In this work, we propose a novel spectral unmixing (SU) strategy using physically motivated parametric endmember (EME) representations to account for temporal spectral variability. By representing the multitemporal mixing process using a state-space formulation, we are able to exploit the Bayesian filtering machinery to estimate the EME variability coefficients. Moreover, by assuming that the temporal variability of the abundances is small over short intervals, an efficient implementation of the expectation–maximization (EM) algorithm is employed to estimate the abundances and the other model parameters. Simulation results indicate that the proposed strategy outperforms state-of-the-art multi-temporal SU (MTSU) algorithms.

Autoencoder (AEC) networks have recently emerged as a promising approach to perform unsupervised hyperspectral unmixing (HU) by associating the latent representations with the abundances, the decoder with the mixing model, and the encoder with its inverse. AECs are especially appealing for nonlinear HU since they lead to unsupervised and model-free algorithms. However, existing approaches fail to explore the fact that the encoder should invert the mixing process, which might reduce their robustness. In this letter, we propose a model-based AEC for nonlinear HU by considering the mixing model a nonlinear fluctuation over a linear mixture. Different from previous works, we show that this restriction naturally imposes a particular structure to both the encoder and decoder networks. This introduces prior information in the AEC without reducing the flexibility of the mixing model. Simulations with synthetic and real data indicate that the proposed strategy improves nonlinear HU.

This contribution presents a novel resampling scheme that leverages Gaussian Processes (GPs) to more accurately approximate the posterior distribution from a set of random measures and, ultimately, enhance resampling by sampling from such approximation. Resampling is a critical step in particle filtering, impacting its estimation performance and parallelization capabilities. The approach can be seen as a kernel-based density approximation. As a byproduct, we are able to i) derive an explicit formula for minimum mean squared error (MMSE) state estimation, and ii) provide a well defined optimization problem for determining the maximum a posteriori (MAP) state estimation. The results on a target tracking problem show the performance improvements of the so-called Gaussian Process Particle Filter (GPPF) when compared to standard particle filtering.

We address a scenario in which networked sensor nodes measure the strength of the field generated by a number of moving objects and transmit their measurements to a fusion center (FC) in a random access manner for the final reconstruction of the objects' trajectories. To ensure scalability over an arbitrary coverage area, we divide the total observation area into design units (DUs), each consisting of several sensing cells. An extended Kalman filter is assigned to each cell, while neighboring cells communicate with each other to exchange current status, thus allowing state fusion whereby the Kalman filters adjust (overwrite) their local estimates at the end of each updating interval. In this manner, provisions are made for the objects leaving and entering a DU, yielding a system design that is scalable across a large coverage area without an increase in complexity (size) of the Kalman filters. We provide a step-by-step procedure for designing the system, taking into account the fact that sensors communicate to the FCs using random access over band-limited and imperfect channels where packet loss is inevitable due to collisions as well as communication noise. In addition, we study different rate control schemes in which the FC instructs the sensors to increase or decrease their sensing (transmission) rate in accordance with the currently estimated object locations. Performance is evaluated through simulation, showing the effectiveness of the approach proposed in terms of the mean squared localization error and data throughput, and quantifying the effect of limited bandwidth and lossy communication.

In this paper we propose a probabilistic framework for proficiency self-assessment of autonomous systems. We define proficiency as a mathematical concept, i.e., as a metric that depends on a variety of factors. This concept allows for assessment of the degree of completion of a given task by a system. We provide the rationale behind the proposed concept and its forms for various settings. Further, we present motivating examples with details of evaluation of the proficiency. We anticipate that our definition of proficiency is a step forward toward achieving "self-awareness" of autonomous systems.

XTS-advanced encryption standard (AES) is an advanced mode of AES for data protection of sector-based devices. It features two secret keys instead of one, and an additional tweak for each data block. These characteristics make the mode not only resistant against cryptoanalysis attacks, but also more challenging for side-channel attack. In this paper, we comprehensively analyze the side-channel power leakage of various XTS-AES implementations and invent effective attacks. We first run a simple power analysis of a software implementation. For a hardware implementation on field-programmable gate array (FPGA), we analyze side-channel leakage of the particular modular multiplication in XTS-AES mode. In addition, we utilize the relationship between two consecutive block tweaks and propose a method to work around the masking of ciphertext by the tweak. These attacks are verified on an FPGA implementation of XTS-AES. The results show that XTS-AES is susceptible to side-channel power analysis attacks, and therefore dedicated protections are required for security of XTS-AES in storage devices.

In localization tasks, one typically assumes a statistical model of the observations, where the model quantifies the observations by exploiting interrelationships based on geometry. These models might incorporate unknown parameters that, in general, are functions of space. In this article, we propose a crowd sensing method for estimating a spatial field of a quantity (e.g., ranging biases due to line-of-sight/non-line-of-sight or path-loss parameter) allowing for improved indoor localization. Our method takes advantage of the information provided by various users that navigate the area of interest. The proposed learning approach is based on Gaussian processes and its computational cost does not increase with the number of measurements. We present numerical results that show how the proposed method estimates a spatial field of biases and how these estimates lead to much improved performance in estimation of user positions.

In this paper, we present a frame synchronization method which consists of the non-orthogonal superposition of a synchronization sequence and the data. We derive the optimum detection criterion and compare it to the classical sequential concatenation of synchronization and data sequences. Computer simulations confirm the benefits of the non-orthogonal allocation for the case of short frames, which makes this technique particularly suited for the increasingly important regime of low latency and ultra- reliable communication.

This special issue addresses various jammers and their effect on different processing stages and overall Global Navigation Satellite System (GNSS) receiver performance, and presents countermeasures and solutions to combat interference.

This paper presents an overview of the coding aspects of a GNSS receiver. Coding allows detection and correction of channel-induced errors at the receiver, here the focus is on the mitigation of threats from malicious interferences. Although the effects of interference at different stages of GNSS baseband processing has been deeply analyzed in the literature, little attention was devoted to its impact on the navigation message decoding stage. Theis paper provides an introduction to the various coding schemes employed by current GNSS signals, discussing their performance in the presence of noise in terms of block-error rate. Additionally, the benefits of soft-decoding schemes for navigation message decoding are highlighted when jamming interferences are present. The proposed scheme requires estimating the noise plus interference power, yielding to enhanced decoding performances under severe jamming conditions. Finally, cryptographic schemes as a means of providing anti-spoofing for geosecurity location-based services, and their potential vulnerability are discussed, with particular emphasis on the dependence on the dependence of the scheme on successful navigation message decoding.

One of the main vulnerabilities of GNSS receivers is their exposure to intentional or unintentional jamming signals, which could even cause service unavailability. Several alternatives to counteract these effects were proposed in the literature, being the most promising those based on multiple antenna architectures. This is specially the case for high-grade receivers used in applications requiring reliability and robustness. This article provides an overview of the possible receiver architectures encompassing antenna arrays and the associated signal processing techniques. Emphasis is also put on the most typical implementation issues found when dealing with such technology. A thorough survey is complemented with a set of experiments, including real data processing by a working prototype, which exemplifies the above ideas.

One of the major challenges in Bayesian filtering is the curse of dimensionality. The quadrature Kalman filter (QKF) is the method of choice in many real-life Gaussian problems, but its computational complexity increases exponentially with the dimension of the state. As a promising solution to overcome the filter limitations in such scenarios, we further explore the multiple state-partitioning approach, which considers the partition of the original space into several subspaces, with the goal to apply a low-dimensional filter at each partition. In this contribution, the key idea is to take advantage of the estimation uncertainty provided by the QKF to improve the interaction among filters and avoid the point estimate approximation performed in the original Multiple QKF (MQKF). The new filter formulation, named Improved MQKF, considers Gauss-Hermite quadrature rules to propagate the subspaces of interest, together with cubature rules for marginalization purposes. The nested quadrature-cubature approximation provides robustness and improves the filter performance. Simulation results for a multiple target tracking scenario are provided to support the discussion.

In this paper we develop two new approaches for frame synchronization in the binary-input AWGN channel, in which we account for the sign ambiguity of the received symbols and exploit knowledge of an alternating sequence which precedes the synchronization word. We present an approach based on an extended sliding window and the appropriate decision metric. For the common case that the synchronization word is followed by encoded data we present a solution which exploits the error detection capability of the channel decoder and applies a list decoding approach for frame synchronization. The proposed methods are validated through computer simulations in the deep-space communication uplink and show significant performance gains compared to current solutions.

Positioning is a key aspect for many applications in wireless sensor networks. In order to design practical positioning algorithms, employment of efficient algorithms that maximize the battery lifetime while achieving a high degree of accuracy is crucial. The number of participating anchor nodes and their transmit power have an important impact on the energy consumption of positoning a node. This paper proposes a game theoretical algorithm to optimize resource usage in obtaining location information in a wireless sensor network. The proposed method provides positioning and tracking of nodes using RSS measurements. We use the Geometric Dilution of Precision as an optimization metric for our algorithm, with the aim of minimizing the number and power of anchor nodes that collaborate in positioning, thus saving energy. The algorithm is shown to be a potential game, therefore convergence is guaranteed. A distributed low complexity solution for the implementation is presented. The game is applied to WSN and results show the trade-off between power saving and positioning error.

This paper derives the Ziv-Zakai lower bound (ZZLB) for the time of arrival (TOA) estimation in the presence of one interfering pulse from which no a priori knowledge is available. The bound is obtained by including the interference in the system model but only the transmitted pulse as a candidate for the likelihood ratio (LR) test. A compact ZZLB expression that depends on the time delay and amplitude of the interference is obtained. We compare the performance of the first path maximum likelihood estimation (MLE) with the bound as a function of the relative distance between the first path and the interfering path.

Predicting the course of influenza rates is extremely useful for the efficacy of planned vaccination programs. In this paper we address this problem by stating a dynamic state-space model that mathematically describes both the evolution of influenza rates and the observations obtained by a surveillance system. We then propose a prediction method based on particle filtering that accommodates the nonlinear nature of the model. Using real data we estimate the necessary model functions prior to the prediction step. Computer simulations reveal promising results of the proposed method.

This paper addresses the problem of inferring voltage traces and ionic channel activity from noisy intracellular recordings in a neuron. A particle filtering method with optimal importance density is proposed to that aim, with the benefits of on-line estimation methods and Bayesian filtering theory. The method is applied to an inaccurate Morris-Lecar neuron model without loss of generality. Simulation results show the validity of the approach, where it is observed that theoretical estimation bounds are attained.

This paper analyzes and validates an enhanced implementation of the multiple particle filter that improves its accuracy when applied to high dimensional problems. The algorithm combines the divide et impera philosophy of the multiple particle filter, which avoids the collapse of traditional particle filters, with game theory strategies that provide with a powerful tool to improve the performance. The problem of multiple target tracking with received signal strength measurements is addressed and the results show remarkable improvement over both standard particle filtering and multiple particle filtering.

This paper investigates the application of the Generalized Likelihood Ratio Test detector to the Global Navigation Satellite System array-based acquisition problem. We consider an unstructured channel model which includes colored noise with an arbitrary covariance matrix. We show that the proposed test function is a Constant False Alarm Rate detector and we provide closed form expressions for false alarm, detection probabilities, and the Receiver Operating Characteristic. Furthermore, this work analyzes the Sample Covariance Matrix structure and the capability of the detector to reject uncorrelated interferences. Simulation results validate the theoretical analysis.

The performance of Global Navigation Satellite Systems (GNSS) is known to be severely affected in multipath scenarios, providing a biased position solution which could jeopardize possible geodetic-grade applications. Indeed, multipath estimation and/or mitigation has attracted the attention of many researchers in the recent years. Nevertheless, few attention has been given to the detection of multipath. That is to say, acknowledging that a given scenario is corrupted by multiple propagation paths, and thus using adequate techniques to combat its effect. This paper proposes a statistical multipath detector based on an antenna array GNSS receiver. Specifically, the detector resorts to the estimated noise covariance matrix to compute a statistic that measures the dispersion of its eigenvalues. The theoretical distribution of such statistic is known. This is used by the proposed method to perform a Kolmogorov-Smirnov one-sample test to assess departures from the theoretical distribution, with the resulting detector being Constant False Alarm Rate. The detector is analyzed in terms of its probability of detection and an analysis is provided regarding its behavior under synchronization errors. Results of the detector under realistic scenarios are also discussed.

In signal processing, it is typical to develop or use a method based on a given model. In practice, however, we almost never know the actual model and we hope that the assumed model is in the neighborhood of the true one. If deviations exist, the method may be more or less sensitive to them. Therefore, it is important to know more about this sensitivity, or in other words, how robust the method is to model deviations. To that end, it is useful to have a metric that can quantify the robustness of the method. In this paper we propose a procedure for developing a variety of metrics for measuring robustness. They are based on a discrete random variable that is generated from observed data and data generated according to past data and the adopted model. This random variable is uniform if the model is correct. When the model deviates from the true one, the distribution of the random variable deviates from the uniform distribution. One can then employ measures for differences between distributions in order to quantify robustness. In this paper we describe the proposed methodology and demonstrate it with simulated data.

This paper considers the problem of ultra-tight GNSS/INS integration. We propose a new approach, deriving the direct relation between Inertial Measurement Unit (IMU) measurements and synchronization parameters, used in the trilateration algorithm to compute the position of the receiver. We take into account the IMU's eventual biased behavior by introducing it into the state representation. We use a recently-developed, square-root derivative-free Gaussian nonlinear filter to solve the estimation problem.

One of the most criticized aspects of particle filtering algorithms is their dependence on model assumptions. However, a rigorous study of the effect of modeling errors on the performance of such algorithms is still missing. In this paper, the problem of using an inaccurate discrete state-space model is considered and a systematic methodology for studying the effects on its performance is proposed. The methodology is based on the use of the Kolmogorov-Smirnov statistic, which in this case is a distance metric between the posterior characterization when respectively correct and incorrect model assumptions are made. An example with functional and distributional inaccuracies is studied.

In this paper, the issue of network topology control in wireless networks using a fully distributed algorithm is considered. Whereas the proposed distributed algorithm is designed applying game theory concepts to design a non-cooperative game, network connectivity is guaranteed based on asymptotic results of network connectivity. Simulations show that for a relatively low node density, the probability that the proposed algorithm leads to a connected network is close to one.

Multipath is one of the dominant sources of error in high-precision GNSS applications. A tracking algorithm is presented that explicitely accounts for direct signal and multipath replicas in the model, in order to mitigate the contributions of the latter. A Bayesian approach has been taken, to infer some information from the time evolution model of the parameters. Due to the nonlinearity of the measurement model, a Particle Filtering algorithm has been designed. The proposed PF considers Rao-Blackwellization with a CKF and the selection of the importance density is performed via the use of Laplace’s method, which yields to an importance density close the optimal. Simulations compare performance to EKF and PCRB.

In this paper, the maximum likelihood estimator (MLE) of the position in satellite based navigation systems is studied. Recent results have shown that this novel approach provides an interesting way of introducing prior information in the position estimation and that the estimator is consistent for large sample sizes. However, one of the main drawbacks of this approach is the lack of a computationally efficient optimization algorithm due to the high dimensionality and nonlinearity of the resulting cost function, since there is not a closed form solution for this estimator. The aim of this paper is to investigate the application of the space-alternating generalized expectation maximization (SAGE) algorithm to the estimation of position. The SAGE algorithm is a low-complexity generalization of the EM (expectation-maximization) algorithm, which iteratively approximates the MLE. Computer simulation results are provided, comparing the performance obtained by the algorithm with the Cramer-Rao bound.

In this letter, we obtain the maximum likelihood estimator of position in the framework of global navigation satellite systems. This theoretical result is the basis of a completely different approach to the positioning problem, in contrast to the conventional two-step position estimation, consisting of estimating the synchronization parameters of the in-view satellites and then performing a position estimation with that information. To the authors' knowledge, this is a novel approach that copes with signal fading, and it mitigates multipath and jamming interferences. Besides, the concept of position-based synchronization is introduced, which states that synchronization parameters can be recovered from a user position estimation. We provide computer simulation results showing the robustness of the proposed approach in fading multipath channels. The root mean square error performance of the proposed algorithm is compared to those achieved with state-of-the-art synchronization techniques. A sequential Monte Carlo-based method is used to deal with the multivariate optimization problem resulting from the maximum likelihood solution in an iterative way

In direct-sequence spread-spectrum(DS-SS) navigation based systems, multipath can degrade seriously synchronization performance causing time delay and code phase estimates, to deviate from the actual value. This bias depends on the relative amplitudes and delays of multipath replicas with respect to the direct signal. The error in the estimated position due to multipath, when using a standard Delay Lock Loop, can be on the order of several tens of meters, which is a critical aspect in high-precision applications. This works presents a Sequential Monte Carlo based algorithm which tries to iteratively estimate complex amplitudes and delays of the direct signal and multipath replicas by characterizing the posterior probability density function of these parameters relying on particle filter theory. Simulations are presented for navigation systems, which are particular applications of DS-SS systems.