Global Navigation Satellite Systems (GNSS) are fundamental for positioning, navigation, and timing (PNT), playing a crucial role in next-generation intelligent transportation systems and safety-critical applications. However, achieving precise PNT solutions in challenging environments remains a significant challenge. Under ideal conditions, carrier-phase-based techniques such as Real-Time Kinematics (RTK) and Precise Point Positioning (PPP) enable high-precision positioning. However, their accuracy heavily depends on the quality of phase observables, which can be degraded in harsh environments, such as urban canyons or interference-prone scenarios. A promising alternative is the use of large-bandwidth signals, which enhance resolution and improve code-based observables. This can be achieved through high-order Binary Offset Carrier modulations or GNSS meta-signals. This study investigates the fundamental performance limits of time delay and Doppler estimation for such signals in challenging scenarios, particularly in the presence of multipath interference, where signal reflections significantly impact receiver performance. Characterizing multipath effects is critical for the next generation of PNT applications, as it directly influences the robustness of GNSS solutions. To analyze these effects, we derive the Cramér-Rao Lower Bound (CRB) for time-delay and Doppler estimation under a signal model where one specular multipath degrades GNSS receiver performance. This case considers that the receiver is aware of the multipath and applies countermeasures. In the second case, we assume that the receiver is unaware of the multipath, for which we derive the Misspecified CRB (MCRB). The MCRB quantifies the performance degradation in standard GNSS receivers due to unmodeled multipath interference. We validate these theoretical bounds by comparing them with state-of-the-art estimation algorithms. Our results demonstrate the significant performance improvements achievable in harsh conditions using metasignals such as Galileo E5a + E5b or GPS L2 CM + L5, compared to legacy signals such as GPS L1 C / A.

The estimation of high-resolution spectrometer instrument spectral response functions (ISRFs) is crucial, for example in the context of remote sensing in order not to compromise the determination of trace gas concentrations. This paper introduces a new statistical model and an optimization algorithm for the joint estimation of ISRFs and spectral shifts. As a key novel ingredient, we investigate the use of optimal transport theory and the associated Wasserstein distance to estimate the spectral shifts, comparing this approach to the conventional ℓ2 norm. As a second key ingredient, a sparse representation of the ISRFs is used to decompose these functions into a fixed dictionary of atoms. Results suggest that the proposed method performs well for small spectral shifts with both distances, while the Wasserstein distance proves particularly effective for estimating larger spectral shifts.

This paper investigates a non-parametric regression approach based on a reproducing kernel Hilbert space framework to model the relationship between Landsat-8 spectral bands and the depth of shallow inland lakes (up to 25m). Unlike existing parametric methods, which rely on predefined assumptions about the relationship between the Landsat bands and the lake depth, the proposed method considers a more flexible non-parametric model based on a radial basis function kernel. This model can handle multiple band ratios to estimate lake depths. The performance of the proposed method is validated on synthetic and real data and compared against traditional parametric models. The results presented in this paper show that the proposed nonparametric model is very competitive in terms of accuracy, while eliminating the need for manual parameter selection, especially in the context of remote sensing of turbid inland water bodies.

This paper presents a Bayesian multifractal segmentation method that segments multifractal textures in regions with different multifractal properties. First, a computationally and statistically efficient model for wavelet leaderbased multifractal parameter estimation is developed, assigning wavelet leader coefficients associated with distinct parameters to different image regions. Next, a multiscale graph label prior is introduced to capture spatial and scale correlations among these labels. Gibbs sampling is used to generate samples from the posterior distribution. Numerical experiments on synthetic multifractal images demonstrate the effectiveness of the proposed method, outperforming traditional unsupervised and modern deep learning-based segmentation approaches.

This paper deals with cyclic redundancy check (CRC) decoding when used in the context of non forward error correction (FEC)-encoded IoT systems: CRC decoding remains challenging when combined to continuous phase modulation (CPM). In this paper, a proposed algorithm relying on the candidate diversity principle through a candidate list generation from soft CPM demodulation output combined with CPM-tailored syndrome decoding is evaluated. Applied with Bahl Cocke Jelinek Raviv (BCJR) algorithm for Gaussian minimum shift keying (GMSK) demodulation, it outperforms all existing complexity-affordable methods and performs close to the best evaluated Parallel-List Viterbi Algorithm with usual CRC validation.

In this work, we investigate the design of the parity-check matrix in the recently introduced Syndrome-Based Neural Decoders, which are powerful neural-network-based channel decoders. We show that the structure of the parity-check matrix has a crucial effect on the decoder’s performance, propose an information-theoretic metric to describe this effect, and suggest an algorithm to construct a parity-check matrix accordingly. The results are illustrated both analytically and numerically, showing a considerable performance gain for the proposed design without any increase in training or decoding complexity.

Cet article présente une extension d’une approche plug-and-play pour le recalage de nuages de points 3D. Le problème de recalage de nuages de points 3D est formulé comme un problème inverse, et une approche plug-and-play est utilisée pour conjointement débruiter et recaler les nuages de points. Dans cet article, nous proposons d’optimiser la transformation de recalage en exploitant la structure de groupe de Lie de la transformation rigide SE(3). Des expériences menées sur des nuages de points LiDAR sont présentées mettant en évidence l’amélioration de la méthode par rapport à une méthode existante.

Différentes méthodes d’imagerie satellitaire sont utilisées pour estimer la bathymétrie des lacs. L’enjeu principal réside dans l’établissement d’une relation précise entre la réflectance dans les bandes spectrales observées et la profondeur de l’eau, une tâche complexe dont les incertitudes peuvent affecter la fiabilité des estimations obtenues. Les approches paramétriques, largement utilisées dans la littérature, reposent sur des modèles établis, tandis que des méthodes non paramétriques ont été explorées plus récemment afin de s’affranchir de certaines hypothèses. Dans cet article, nous proposons une approche bayésienne fondée sur les processus Gaussiens, permettant une modélisation probabiliste des relations spectro-bathymétriques ainsi qu’une quantification rigoureuse des incertitudes associées aux estimations de profondeur.

Cet article présente un nouvel algorithme EM (Expectation-Maximization) pour le recalage robuste de nuages de points 2D–3D issus d’une caméra et d’une carte de référence. Nous nous intéressons à l’estimation conjointe des paramètres d’intérêt (i.e., orientation et position de la caméra), de la proportion d’observations aberrantes et de la variance du bruit de mesure. L’approche proposée repose sur un modèle statistique intégrant des variables latentes permettant de gérer les associations inconnues entre points 2D, points 3D et observations aberrantes, via un modèle de mélange. Des résultats obtenus à partir de données synthétiques montrent l’intérêt de cette démarche en termes de rapidité de convergence de l’algorithme proposé et de robustesse face aux mesures aberrantes.

Les observations de phase de la porteuse dans les récepteurs GNSS permettent un positionnement au centimètre près mais sont affectées par un bruit de phase supposé qui suit une distribution de von Mises, dégradant la performance des estimateurs. Nous proposons une approche novatrice contraignant les paramètres de von Mises—localisation angulaire et dispersion—dans l’espace du groupe de Lie SO(2) × R+. Un estimateur du maximum de vraisemblance sur groupes de Lie, résolu via un algorithme de Newton, améliore la rigueur mathématique et la précision, notamment avec peu d’observations, par rapport aux méthodes euclidiennes.