Packet loss due to misrouted or delayed packets in voice over IP leads to huge voice quality degradation. Packet loss con- cealment algorithms try to enhance the quality of the speech. This paper presents a new packet loss concealment algorithm which relies on one hidden Markov model. For this purpose, we introduce a continuous observation vector well-suited for silence, voiced and unvoiced sounds. We show that having a global HMM is relevant for this application. The proposed system is evaluated using standard PESQ score in a real- world application.

SpaceWire is a standard for on-board satellite networks chosen by the ESA as the basis for future data-handling architectures. However, network designers need tools to ensure that the network is able to deliver critical messages on time. Current research only seek to determine probabilistic results for end-to-end delays on wormhole networks like SpaceWire. This does not provide sufficient guarantee for critical traffic. Thus, in this paper, we propose a method to compute an upper-bound on the worst-case end-to-end delay of a packet in a SpaceWire network.

In automatic speech analysis, voicing articulatory is often defined as a binary decision: voiced or unvoiced. Lin- guists agree that this articulatory should be continuous. In this paper, we present a new approach to compute a con- tinuous voicing indicator of a speech frame. This voic- ing percentage is then evaluated in both a segmentation process and a speech recognition task. Promising results show that this continuous voicing percentage may be used as a reliable voicing indicator.

Detecting the presence of non-stationarity events in a signal is a challenge that is still not taken up. The aim of this paper is to make a contribution to this key issue. We already proposed a non-stationarity detection defined in time-frequency domain in order to control the invariance of the time-frequency statistics. In this paper, in order to be not limited by the time and frequency resolution of a time-frequency approach, we propose another test in frequency domain. In frequency domain, the problem can be cast by taking advantage of the normalized-variance properties of a spectral estimator when analyzing non-stationary signals. This second test will confirm, invalidate or detect new frequency localizations of non-stationarities. Finally, the main contribution of the paper is to propose a stationary index defined so as to merge the information given by these two tests and to allow an alarm to be raised for a high level of non-stationarities. Applications on real-world signals show the pertinence of this new index.

Satellite navigation has acquired an increased importance during these last years, on the one hand due to the imminent appearance of the European GALILEO system that will complement the American GPS, and on the other hand due to the great success it has encountered in the commercial civil market. An important part of this success is based on the technological development at the receiver level that has rendered satellite navigation possible even in difficult environments. Today’s objective is to prepare the utilisation of this kind of signals for land vehicle applications demanding high precision positioning. One of the main challenges within this research domain, which cannot be addressed by classical coupling techniques, is related to the system capability to provide reliable position estimations. The enhancement in dead-reckoning technologies (i.e. size reduction of MEMS-based sensors or gyroscopes) cannot all by itself reach the necessary confidence levels if exploited with classical localization and integration algorithms. Indeed, these techniques provide a position estimation whose reliability or confidence level it is very difficult to quantify. The feasibility of these applications relies not only on an extensive research to enhance the navigation algorithm performances in harsh scenarios, but also and in parallel, on the possibility to maintain, thanks to the presence of additional sensors, a high confidence level on the position estimation even in the absence of satellite navigation signals.

This thesis studies the applicability of OFDM techniques for future satellite telecommunications systems. In particular, he treats the synchronization problem at the receiver for satellite broadcasting systems in Ka band. The system uses QAM modulation with M phase stages and works in continuous mode, at small signal to noise ratio (typically at Eb/N0 = 0dB). The main objective of this thesis is to propose a receiver synchronization structure using as least resources as possible in order to optimize spectral efficiency. Two studies are carried out. The first study consists of proposing and validating a synchronization structure in the aim of optimizing spectral efficiency. The second study evaluates the performance of this structure then, compares it with existing standards such as DVB-S and DVB-S2 in terms of spectral efficiency. For first study, synchronization errors have been identified and their impact on system performance evaluated. These results prove that excepting clock frequency error, other synchronization errors have to be estimed and corrected. The transmission in continuous mode for fixed satellite broadcasting system allows the use of NDA (Non-Data-Aided) loop structure in order to avoid the use of pilots, then improve spectral efficiency. However, these algorithms need a coarse synchronization stage in order to limit interference terms. So, the synchronization includes a coarse stage in order to limit interference terms and a finer stage in order to improve system performance. For coarse synchronization stage, simulation results prove that algorithms using guard interval give better performance than whom using pilots. The second study evaluates performance of the proposed structure. In coarse synchronization stage, this evaluation allows to specify guard interval length then, calculate and compare system performance in terms of spectral efficiency with its in DVB-S and DVB-S2 standards. Then, performance evaluation of fine synchronization stage allows to calculate degradations of the proposed structure in the absence and in presence of phase noise models, one of important parameters in a telecommunications system. This stage allows not only to define operating points of the proposed structure in presence of phase noise models in existing standards but also to define phase noise mask of Wiener model supported by this structure. Finally, a study on the hanging time of the proposed structure allows to evaluate the time neccesary to this structure, based on blind algorithms, to be converged. Another study also evaluates the complexity of this structure. This study shows that the synchronization structure proposed in this thesis uses little resources both in terms of spectral efficiency and number of calculations.

This paper presents the modelling of new WLAN models for different indoor environments. This work was carried out in the frame of the FIL project which is funded by the French research agency ANR, in collaboration with Thales Alenia Space France. Based on the standard Opnet models for WLAN nodes, the propagation loss estimation for these types of environment has been improved. We derive an empirical model for spatial registration patterns of mobile users as they move within a TeSA Labs wireless local area network (WLAN) environment and register signal power from different access points. Such a model can be very useful in a variety of simulation studies of the performance of mobile wireless systems, to address issues such as resource management and mobility management protocols. We base the model on extensive experimental data from a TeSA Labs 2.4 GHz WiFi LAN installation. We divide the empirical data available to us into training and test data sets, develop the model based on the training set, and evaluate it against the test set. The new scenarios used to simulate these new propagation models are shown. Finally, results, conclusions and further work are given.

Hyperspectral images present some specific characteristics that should be used by an efficient compression system. In compression, wavelets have shown a good adaptability to a wide range of data, while being of reasonable complexity. Some wavelet-based compression algorithms have been successfully used for some hyperspectral space missions. This paper focuses on the optimization of a full wavelet compression system for hyperspectral images. Each step of the compression algorithm is studied and optimized. First, an algorithm to find the optimal 3-D wavelet decomposition in a rate-distortion sense is defined. Then, it is shown that a specific fixed decomposition has almost the same performance, while being more useful in terms of complexity issues. It is shown that this decomposition significantly improves the classical isotropic decomposition. One of the most useful properties of this fixed decomposition is that it allows the use of zero tree algorithms. Various tree structures, creating a relationship between coefficients, are compared. Two efficient compression methods based on zerotree coding (EZW and SPIHT) are adapted on this near-optimal decomposition with the best tree structure found. Performances are compared with the adaptation of JPEG 2000 for hyperspectral images on six different areas presenting different statistical properties.

The French Space Agency, CNES, is interested in the transforms derived from the wavelets in order to increase the image compression efficiency on-board of Earth observation satellites. In this thesis, the post-transforms are studied. They are employed after the wavelet transform. Each block of wavelet coefficients is further transformed in a basis selected among a dictionary by minimization of a rate-distortion criterion. First, we emphasize dependencies between wavelet coefficients limiting the compression efficiency. Then, we study the bandelet transform by blocks, from which the post-transforms derive, and we optimize its parameter for the compression of satellite images. Particularly, we adapt Shoham and Gersho optimization method to the problem of the selection of the best bandelet basis. We deduce from these results an expression of the optimal Lagrangian multiplier used in the rate-distortion criterion. Next, we analyze dependencies between wavelet coefficient which are not exploited by the bandelet transform and we define new post-transform bases. Bases build by PCA minimize the correlations between post-transformed coefficients and compact the energy of each block on a small number of coefficients. This feature is exploited during the entropy coding process. Last, we modify the bases selection criterion to adapt the post-transform to progressive compression schemes. We then employ the Hadamard post-transform with the CCSDS image encoder to obtain a low computational complexity yet efficient compression scheme.

Satellite networks have been built by the DVB group and dedicated to digital television service. However, in the current service convergence trend, a future satellite network architecture has to be built in a less dedicated way to fit heterogeneous services. This work begins with a description of DVB satellite networks. Then, network convergence solutions are studied for the satellite context. IP and MPLS have then been chosen to build a satellite convergent architecture. Several scenarios are examined so as to evaluate this architecture. A first one deals with the historical television service in a unidirectional, transparent satellite context. We show the feasibility of such a scenario with similar performances and better protocol organisation which simplifies satellite evolution. The next scenarios concern an interactive television service with a return link and a voice over IP service. The ability of deploying new services in a simple manner is thus highlighted. The last scenario applies our convergent approach to the ULISS industrial project of a regenerative hybrid satellite. It shows the flexibility of our architecture and expand ULISS service capabilities.