Cooperative systems used for vessel identification and localization in the context of maritime surveillance, such as the Automatic Identification System (AIS), are often coupled to systems that allow the observation of uncooperative ships such as the Synthetic Aperture Radar (SAR). The combination of information coming from the SAR image and AIS signals can improve the detection of some ships in dense environments, but also allows possible piracy scenarios to be identified. The most common approach for data fusion is the “fusion after detection”, where each system processes the raw data independently. In the context of AIS and Radar, three levels of fusion can be identified: 1) fusion of the raw data, 2) fusion of raw data from a system with the processed data (list of detection) from the other system, 3) fusion of the detection lists formed by the two systems. We will focus on the first two cases, since the last case has been more widely covered in the literature. After introducing the AIS and Radar systems for maritime surveillance, we present structure of AIS data and radar signals, as well as the signal processing used to decode these AIS signals or to produce a radar image. The second chapter presents the potential benefits of the joint use of raw data from both radar and AIS for ship detection. After having described the signal models associated with the unknown ship position, we investigate the detection problem using a Generalized Likelihood Ratio Test (GLRT). The theoretical performances of this test are evaluated and allow us to estimate the performance gain in comparison to a single RSO processing. These theoretical results are validated by Monte Carlo simulations using Receiver Operational Characteristics (ROC). The detection results obtained using the GLRT are encouraging. However, the time implementation of these methods for practical applications is complicated. We therefore proceed to a sub-optimal detector using raw data from the radar and a list of detections from the AIS system, leading to a more simple detection strategy. The third chapter studies the fusion of raw radar data with a list of ship positions, formerly provided by the AIS system. Since the ships are moving and the AIS and Radar measurements are not are not acquired at the same time instants, the ship positions have to be extrapolated. Two extrapolation cases are considered in this work: 1) extrapolation errors are lower than the radar resolution and do not have to be integrated in the model, 2) extrapolation errors are not negligible and have to be taken into account in the model. Contrary to the second chapter, four hypotheses can now be considered. Indeed, in addition to the classical detection scenarios by both systems, we can identify the cases where only one of the systems detects a ship, which corresponds to the situations where a ship does not transmit its AIS position or where a ship intentionally false its AIS position. The problem can then be formalized with two successive binary hypothesis tests. This process allows the information coming from AIS and radar data to be fused naturally, aleading to improved radar detection performance. A performance comparison of this detector that uses a priori information with conventional radar detection shows that it is less sensitive to the proximity to other ships and to the ship density of the considered scenario. The fourth chapter presents the signal simulator considered in this thesis to test the detection algorithms in different surveillance scenarios, i.e., a piracy ship hijacking scenario, an illegal cargo transshipment and a navigation in a dense environment.

Cooperative systems used for vessel identification and localization in the context of maritime surveillance, such as the Automatic Identification System (AIS), are often coupled to systems that allow the observation of uncooperative ships such as the Synthetic Aperture Radar (SAR). The combination of information coming from the SAR image and AIS signals can improve the detection of some ships in dense environments, but also allows possible piracy scenarios to be identified. The most common approach for data fusion is the “fusion after detection”, where each system processes the raw data independently. In the context of AIS and Radar, three levels of fusion can be identified: 1) fusion of the raw data, 2) fusion of raw data from a system with the processed data (list of detection) from the other system, 3) fusion of the detection lists formed by the two systems. We will focus on the first two cases, since the last case has been more widely covered in the literature. After introducing the AIS and Radar systems for maritime surveillance, we present structure of AIS data and radar signals, as well as the signal processing used to decode these AIS signals or to produce a radar image. The second chapter presents the potential benefits of the joint use of raw data from both radar and AIS for ship detection. After having described the signal models associated with the unknown ship position, we investigate the detection problem using a Generalized Likelihood Ratio Test (GLRT). The theoretical performances of this test are evaluated and allow us to estimate the performance gain in comparison to a single RSO processing. These theoretical results are validated by Monte Carlo simulations using Receiver Operational Characteristics (ROC). The detection results obtained using the GLRT are encouraging. However, the time implementation of these methods for practical applications is complicated. We therefore proceed to a sub-optimal detector using raw data from the radar and a list of detections from the AIS system, leading to a more simple detection strategy. The third chapter studies the fusion of raw radar data with a list of ship positions, formerly provided by the AIS system. Since the ships are moving and the AIS and Radar measurements are not are not acquired at the same time instants, the ship positions have to be extrapolated. Two extrapolation cases are considered in this work: 1) extrapolation errors are lower than the radar resolution and do not have to be integrated in the model, 2) extrapolation errors are not negligible and have to be taken into account in the model. Contrary to the second chapter, four hypotheses can now be considered. Indeed, in addition to the classical detection scenarios by both systems, we can identify the cases where only one of the systems detects a ship, which corresponds to the situations where a ship does not transmit its AIS position or where a ship intentionally false its AIS position. The problem can then be formalized with two successive binary hypothesis tests. This process allows the information coming from AIS and radar data to be fused naturally, aleading to improved radar detection performance. A performance comparison of this detector that uses a priori information with conventional radar detection shows that it is less sensitive to the proximity to other ships and to the ship density of the considered scenario. The fourth chapter presents the signal simulator considered in this thesis to test the detection algorithms in different surveillance scenarios, i.e., a piracy ship hijacking scenario, an illegal cargo transshipment and a navigation in a dense environment.

Recent trends in Global Navigation Satellite System (GNSS) applications in urban environments have led to a proliferation of research works that seek to mitigate the adverse effect of Multipaths (MPs) and non-line-of-sight (NLOS). For such harsh urban settings, this paper proposes an original methodology for constructive use of degraded MP/NLOS signals, instead of their elimination, via a fusion of GNSS pseudoranges (PR) with aided information from a 3D GNSS simulator. First, a 3D GNSS simulator is used to characterize and predict PR measurements over an array of candidate positions in the environment under study. Then, a similarity scoring technique based on least-squares (LS) position matching is applied to score candidate positions. Finally, the final position estimate is retained as the weighted average of the candidate positions with the highest scores. Experiment results using real GNSS data in a deep urban environment confirm the theoretical sub-optimal efficiency of the proposed approach, despite it intensive computational load.

The goal of this position paper is to contribute for the improving of Cyber-Physical System (CPS) simulations by introducing distribution. CPS use computations andcommunication tightly interacting with physical processes. So a CPS simulation needs to tackle with three kinds of simulations: the computational simulation, the physical simulation, and the communication simulation. In this paper, we will focus on the communication simulation, and its interaction with the two others simulations. We will draw a landscape of the existing concepts and technologies for distributing communication simulation, and then propose an architecture for interacting with the whole CPS simulation. We will apply this architecture to a simulation of a satellite constellation, where satellites can be simulated with different levels of precision, from the simple generic mathematical model to the heavy-featured CPS simulation.

Performance estimation of coded LEO satellite-to ground laser transmissions partially corrected by adaptive-optics are presented. Through numerical simulations, the conjugation of adaptive-optics with a cross-layering optimization of data reliability mechanisms is investigated. The emphasis is put on the minimization of the data memories required at both the transmitter and the receiver in order to guarantee an error-free downlink.

In this paper, we present a generalized polyphase representation for Continuous Phase Modulation (CPM) signals suited to the detection over frequency-selective channels. We first develop two different equalizers based on this representation and relate them to the State of Art. We also derive a Least Squares (LS) channel estimation and an improved LS estimation using a priori on the channel. Simulation results show the equivalence between existing equalizers and also show that our channel estimation leads only to a small degradation in term of Bit Error Rate (BER) in the case of an aeronautical communication over a satellite link.

Spectral analysis of sea level anomalies (SLA) is widely used in the altimetry community to understand the geophysical content of the measured signal, to assess and compare the outputs of different missions. Spectral content of SLA is used to characterize ocean at different scales and to estimate the instrumental noise. Based on the SLA spectrum, one can estimate the spectral slope at medium to large scales (relied to the Surface Quasi-Geostrophic (SQG) ocean dynamics theory) and the measurement noise (observed as a noise plateau at smallest scales). A previous contribution [1] has pointed out the weaknesses of spectral analysis based on Fourier transform, mainly due to : (1) the convolutive bias which results in a biased estimation of the slope, the bias being related to the kind of observation weighting temporal window used and (2) the high variance of estimation leading to averaging several spectral estimations and raising the question of stationarity. To overcome these two drawbacks, a parametric spectral analysis method is proposed. This method is based on Auto-Regressive (AR) modeling [2,3] which is known to provide a spectral estimation with a lower variance than the as outperforming Fourier-based methods in terms of variance, in the case of short observation windows, without any need for choosing a weighting temporal window. Moreover, in order to better match the SLA frequency contents on a log scale to match the log scale interest of the SLA frequency contents , warping is introduced as a preprocessing prior to spectral analysis as it is done in speech coding [4]. Comparisons between the proposed parametric method (called ARWARP) and classical Fourier Fourier-based methods have been performed on both simulated SLA signals obtained from theoretical spectra and real signals from a high-resolution altimeter SARAL/AltiKa at 40 Hz rate (Orbit – Range – Mean Sea Surface). Results on simulated SLA signals highlight the performance of the ARWARPmethod, in terms of bias and variance on spectral estimation. ARWARP can be applied on short segments of SLA signals, providing a local information of the ocean characteristics, which can be of promising use by the wider Cal/Val and altimetry science community.

La navigation par satellites tient une place de plus en plus importante dans notre vie. On peut penser par exemple aux véhicules autonomes ou aux péages par localisation, mais plus quotidiennement on peut remarquer que, là où il y a un peu moins de 10 ans on avait besoin d’acheter un récepteur GPS pour naviguer par satellite, aujourd’hui un simple smartphone suffit, proposant de surcroît une plus grande diversité d’utilisation (navigation piétonne, localisation d’objets perdus, Pokémon GO…). Paradoxalement, ces dernières utilisations se font la plupart du temps dans le scénario de navigation le plus complexe : l’environnement urbain. En effet, si l’on arrive plus ou moins bien à estimer la plupart des différents types d’erreurs de propagation des signaux satellites que nous présenterons, les multitrajets restent la source d’erreur la plus compliquée à gérer car très dépendante du scénario : géométrie et surfaces des bâtiments, obstacles temporaires ou feuillage des arbres par exemple. Nous proposerons une solution innovante qui consiste à traiter les erreurs de multitrajets comme des vecteurs de biais dont certaines composantes sont nulles (c.-à-d., nous supposons que certains canaux de propagation ne souffrent pas de multitrajet) et nous appliquerons donc des méthodes d’estimation dite « parcimonieuses ».

L’accès à Internet par satellite permet de relier certains lieux isolés, où le déploiement d’un réseau terrestre est impossible. L’utilisation de constellations de satellites pose cependant des problèmes spécifiques à ce type de réseau, comme des délais longs et variables, ou des forts taux d’erreurs, notamment entre le dernier satellite sur le chemin du message et le récepteur mobile au sol (canal LMS, Land Mobile Satellite). Pour palier ce canal difficile, des mécanismes de fiabilisation ont été introduits au niveau des couches liaison et physique. Ces mécanismes allient l’utilisation de codes FEC et des retransmissions afin d’optimiser l’utilisation du canal LMS. En ce qui concerne la couche transport, le protocole le plus couramment utilisé est TCP, mais ce dernier fonctionne très mal avec les mécanismes de fiabilisation du canal LMS. Lors de cette présentation, nous verrons pourquoi les performances de TCP s’effondrent dans le cas de communications par satellite, puis nous verrons les solutions proposées dans cette thèse pour améliorer les performances de TCP.

This paper aims at quantifying the impact of a default TCP option, known as Delayed Acknowledgment (DelAck), in the context of LEO satellite constellations. Satellite transmissions can suffer from high channel impairments, especially on the link between a satellite and a ground gateway. To cope with these errors, physical and link layer reliability schemes have been introduced, at the price of an increase of the end-to-end delay seen by the transport layer (e.g. TCP). Although DelAck is used to decrease the feedback path load and for overall system performance, the use of this option conjointly with satellite link layer recovery schemes might increase the delay and might be counterproductive. To assess the impact of this option, we drive simulation measurements with two well-deployed TCP variants. The results show that the performance gain depends on the variant used and that this option should be carefully set or disabled as a function of the network characteristics. DelAck has a negative impact on TCP variants which are more aggressive such as TCP Hybla, and should be disabled for these versions. However, it shows benefits for TCP variants less aggressive such as NewReno.