The delineation of P and T waves is important for the interpretation of ECG signals. In this work, we propose a sequential Bayesian detection-estimation algorithm for simultaneous P and T wave detection, delineation, and waveform estimation on a beat-to-beat basis. Our method is based on a dynamic model which exploits the sequential nature of the ECG by introducing a random walk model to the waveforms. The core of the method is a marginalized particle filter that efficiently resolves the unknown parameters of the dynamic model. The proposed algorithm is evaluated on the annotated QT database and compared with state-of-the-art methods. Its on-line characteristic is ideally suited for real-time ECG monitoring and arrhythmia analysis.

Global aircraft optimization is a main concern for future and upcoming programs. In particular, great research efforts are dedicated to Electrical Flight Control Systems (EFCS). Obviously, their reliability increases with the redundancy of the flight parameter sensors. However, physical redundancy, obtained by increasing the number of sensors, penalizes the aircraft weight and cost. This paper proposes a sensor failure detection method based on analytic redundancy. The flight parameter of interest is modelled as a linear function of independent sensor measurements on a sliding observation window. The Partial Least Squares (PLS) algorithm is used to estimate regression coefficients on this window. The PLS computes the solution via an iterative processing, and thus can be implemented in the flight control computer for a real time use. Two different failure detection strategies based on the behaviour of the regression coefficients are proposed. Simulation results show that the proposed method leads to robust detections.

In numerous applications, it is required to estimate the principal subspace of the data, possibly from a very limited number of samples. Additionally, it often occurs that some rough knowledge about this subspace is available and could be used to improve subspace estimation accuracy in this case. This is the problem we address herein and, in order to solve it, a Bayesian approach is proposed. The main idea consists of using the CS decomposition of the semi-orthogonal matrix whose columns span the subspace of interest. This parametrization is intuitively appealing and allows for non informative prior distributions of the matrices involved in the CS decomposition and very mild assumptions about the angles between the actual subspace and the prior subspace. The posterior distributions are derived and a Gibbs sampling scheme is presented to obtain the minimum mean-square distance estimator of the subspace of interest. Numerical simulations and an application to real hyperspectral data assess the validity and the performances of the estimator.

This paper addresses the problem of jointly estimating the statistical distribution and segmenting lesions in multiple-tissue high-frequency skin ultrasound images. The distribution of multiple-tissue images is modeled as a spatially coherent Þnite mixture of heavy-tailed Rayleigh distributions. Spatial coherence inherent to biological tissues is modeled by enforcing local dependence between the mixture components. An original Bayesian algorithm combined with a Markov chain Monte Carlo method is then proposed to jointly estimate the mixture parameters and a label-vector associating each voxel to a tissue. More precisely, a hybrid Metropolis-within-Gibbs sampler is used to draw samples that are asymptotically distributed according to the posterior distribution of the Bayesian model. The Bayesian estimators of the model parameters are then computed from the generated samples. Simulation results are conducted on synthetic data to illustrate the performance of the proposed estimation strategy. The method is then successfully applied to the segmentation of in vivo skin tumors in high-frequency 2-D and 3-D ultrasound images.

Road maps are important for several applications in vehicular networks. As a consequence, an updated map is essential for a proper behavior of such applications. In this paper, we focus on a map update application based on an infrastructure-to-vehicle communication with a high mobility. In order to understand the application's behavior, we analyze different performance criteria: System Goodput, Packet Delivery Ratio, Delay, Fairness, and Fragmentation. We compare the application's QoS behavior with three different flow densities (overloaded system, maximum system capacity, and under saturated system) and determinate a trade off between bandwidth (spectrum efficiency) and performance.

Here the concept of Doppler information geometry is summarized and introduced to evaluate the richness of Doppler velocity components of radar signal and applied for wake turbulence monitoring. With the methods of information geometry, we consider all the Toeplitz Hermitian Positive Definite covariance matrices of order n as a manifold Ωn . Geometries of covariance matrices based on two kinds of radar data models are presented. Finally, a radar detector based on Doppler entropy assessment is analyzed and applied for wake turbulence monitoring. This advanced Doppler processing chain is also implemented by CUDA codes for GPU parallel computation.

Space Wire is a standard of on-board networks for satellites promoted by the ESA. As the ESA plans to use Space Wire as the sole network for both critical and non-critical traffics, network designers need tools to check that all the critical messages meet their deadlines. We previously proposed two such tools to compute an upper-bound on the worst-case end-to-end delay of a packet traversing a Space Wire network. The main contribution of this paper is the comparison of those two methods on a network configuration provided by Thales Alenia Space that is representative of next generation large satellites. The goal is to identify the key parameters that affect the bounds computed by the methods. We then conduct a sensitivity analysis on simpler network configurations to study the impact of those parameters on the methods and determine which method works better in different situations.

In this paper, we introduce a GNSS positioning approach that uses constructively non-line-of-sight (NLOS) signals. A 3D model of the environment is used to predict the geometric paths of NLOS signals. More precisely, we propose a version of the extended Kalman filter augmented by a 3D model, referred to as 3D AEKF, for GNSS navigation in NLOS context. In the proposed approach, the measurement model traditionally based on the trilateration equations is constructed from the received paths estimated by the 3D model. The Jacobian of the measurement model is calculated through knowledge of the wall on which the reflection has occured. To use even less reliable measurements, a robust version of the 3D AEKF is also proposed. Simulations conducted in different realistic configurations allow the performance of the proposed method to be evaluated.

Global Navigation Satellite Systems (GNSS) integrity is defined as a measure of the trust that can be placed in the correctness of the information supplied by the navigation system. Although the concept of GNSS integrity has been originally developed in the civil aviation framework as part of the International Civil Aviation Organization (ICAO) requirements for using GNSS in the Communications, Navigation, and Surveillance / Air Traffic Management (CNS/ATM) system, a wide range of non-aviation applications need reliable GNSS navigation with integrity, many of them in urban environments. GNSS integrity monitoring is a key component in Safety of Life (SoL) applications such as aviation, and in the so-called liability critical applications like GNSS-based electronic toll collection, in which positioning errors may have negative legal or economic consequences. At present, GPS integrity monitoring relies on different augmentation systems (GBAS, SBAS, ABAS) that have been conceived to meet the ICAO requirements in civil aviation operations. For this reason, the use of integrity monitoring techniques and systems inherited from civil aviation in non-aviation applications needs to be analyzed, especially in urban environments, which are frequently more challenging than typical aviation environments. Each application has its own requirements and constraints, so the most suitable integrity monitoring technique varies from one application to another. This work focuses on Electronic Toll Collection (ETC) systems based on GNSS in urban environments. Satellite navigation is one of the technologies the EU recommends for the European Electronic Toll Service (EETS), and it is already being adopted: as of 2012, toll systems for freight transport that use GPS as primary technology are operational in Germany and Slovakia, and France envisages to establish a similar system from 2013. This dissertation begins presenting first the concept of integrity in civil aviation in order to understand the objectives and constraints of existing GNSS integrity monitoring systems. The derivation of the GNSS integrity requirements and the appropriate integrity monitoring techniques capable to meet them needs a deep knowledge of the targeted application and of its needs and constraints. Consequently, a thorough analysis of GNSS-based ETC systems and of GNSS navigation in urban environments is done in Chapter 2 with the aim of identifying the most suitable road toll schemes, GNSS receiver configurations and integrity monitoring mechanisms. In this case, the need of integrity is not given by safety reasons as in civil aviation, but rather as requirements of economic nature (overcharging and undercharging). Geo-fencing is selected as the method for developing GNSS-based ETC systems over a given area or road network, dividing the tolled region in geo-objects which are the basic pricing sections. A simple user detection (geo-object recognition) algorithm is proposed to charge a user the price of a section whenever it is detected inside it. Receiver autonomous integrity monitoring (RAIM) is chosen among other integrity monitoring systems due to its design flexibility and adaptability to urban environments. Finally, the most promising GNSS receivers are retained, giving special emphasis to dual constellation GPS & Galileo users. The use of SBAS corrections is optionally considered for reducing pseudorange errors. An accurate pseudorange measurement model is a key input in the derivation of the GNSS integrity requirements and the evaluation of RAIM performance. A nominal pseudorange measurement model suitable for integrity-driven applications in urban environments has been obtained following a methodology similar to that of civil aviation, splitting the total pseudorange error into five independent error sources which can be modeled independently: broadcasted satellite clock corrections and ephemeris errors, ionospheric delay, tropospheric delay, receiver thermal noise (plus interferences) and multipath. Nominal errors are modeled as zero-mean Gaussian variables, which is consistent with the error models and integrity monitoring systems used in civil aviation, as well as with SBAS corrections. In this work the fault model that includes all non-nominal errors consists only of major service failures. Once the ETC scheme and the fault model are known, the GNSS integrity requirements with which design the RAIM can be calculated. First, the top level requirements of toll applications are defined in terms of maximum allowable probabilities of missed and false geo-object recognition. Afterwards, the relationship between positioning failures and incorrect geo-object recognition is studied, resulting in maximum allowed probabilities of missed and false alarm values that depend on the number of independent positions with available RAIM employed to decide whether the user is or not inside a pricing section. Two RAIM algorithms are studied. The first of them is the Weighted Least Squares Residual (WLSR) RAIM, widely used in civil aviation and usually set as the reference against which other RAIM techniques are compared. Since one of the main challenges of RAIM algorithms in urban environments is the high unavailability rate because of the bad user/satellite geometry, a new RAIM is proposed. The novel algorithm, based on the WLSR RAIM, is designed with the premise of providing a trade-off between the false alarm probability and the RAIM availability in order to maximize the probability that the RAIM declares valid a fault-free position. Finally, simulations have been carried out to study the performance of the different RAIM and ETC systems in rural and urban environments. Electronic toll collection by means of GNSS in urban environments with geo-objects shorter than 500 m and road topologies that allow a Horizontal Alert Limit (HAL) of 25 or 50 meters has been demonstrated to be feasible with certain signal combinations of dual constellation GPS & Galileo users. Single constellation users only attain the requirements in rural environments with some receiver configurations. In all cases, the availability obtained with the novel RAIM improve those of the standard WLSR RAIM. The main contributions of this thesis are a detailed analysis of GNSS-based ETC systems, a numerical pseudorange nominal error model due to ionospheric delay in Galileo single-frequency receivers, a pseudorange nominal error model due to multipath in urban environments suitable for applications with GNSS integrity, the failure tree that leads to geo-object misleading positions, the derivation of the and of fault detection RAIM algorithms for GNSS-based ETC in the case of a threat model consisting on major service failures, a novel RAIM, based on the WLSR RAIM, that increments the number of valid positions within the integrity requirements in urban environments, the derivation of the analytical expression of the chi-squared non-centrality parameter in the WLSR RAIM and the derivation of the null correlation between the test statistic and the navigation solution error in the WLSR RAIM.

The subject of this thesis is to study Bayesian estimation/detection algorithms suitable for P and T wave analysis in ECG signals. In this work, different statistical models and associated Bayesian methods are proposed to solve simultaneously the P and T wave delineation task (determination of the positions of the peaks and boundaries of the individual waves) and the waveform-estimation problem. These models take into account appropriate prior distributions for the unknown parameters (wave locations and amplitudes, and waveform coefficients). These prior distributions are combined with the likelihood of the observed data to provide the posterior distribution of the unknown parameters. Due to the complexity of the resulting posterior distributions, we propose to use Markov chain Monte Carlo algorithms for (sample-based) detection/estimation.