L’égalisation numérique en communications

This tutorial contribution explains how digital equalization permits very high transmission rates, even with severe channels, by adaptive (real time) correction of the distortion. Equalizers are in general digital filters. Transversal equalizers are only suitable for mild channels. In data transmission, where symbols have discrete levels, severe channels can be equalized by adding a recursive path that is filled in with detected symbols. Engineers have realized optimal equalizers for more than twenty years thanks to the adaptive tracking of the channel time variations, at a very low computational cost. However adaptation requires the periodic transmission of a training sequence deprived of information content. This supervised learning technique is acceptable only in an end-to-end communication system. The most recent equalization methods are usable in multiuser systems such as networks, broadcasted communications, etc., because they employ self-learning or unsupervised equalization. Then adaptation is controlled by the very information data flow. It only takes advantage of an a priori statistical knowledge on the emitted data, e.g. their whiteness, a property that is ensured thanks to jamming. Let us conjecture that all equalizers will be self-learning in a near future.     

Statistical model of an undermoded reverberation chamber

Weibull distribution is adopted to model the electric field component of a Reverberation Chamber (RC). Its first property is to include the asymptotic laws, such as Rayleigh and exponential, and its main advantage lies in the fact that the Weibull shape parameter enables a model of the departure from overmoded to undermoded RC regime. Applications are given, such as an RC modal finite element modeling and a Monte Carlo simulation: they prove that the Weibull two-parameter distribution correctly models the quality factor influence. Moreover, the relevance of the use of this extreme value distribution is illustrated.     

Comportement chaotique des filtres récursifs adaptatifs: application à la compression des signaux téléphoniques

Adaptive filtering gives simple iterative methods for extracting a useful signal by linearly filtering an observation correlated with this useful signal. When the observation contains past values of the filter output—as is the case for a recursive structure— the adaptive filter is a non linear operator which can generate complex behaviours even though adaptation is a standard derivative of the gradient algorithm. Such behaviours are observed in a particular case : theArma adaptive prediction with a sinusoidal input. A slow adaptation speed involves a quasi-periodic behaviour due to the locally unstable character of the predictor. This phenomenon called selfstabilization persists as the adaptation speed increases. Its study is more complicated : the behaviour becomes chaotic. As an illustration, the digitization of telephone signals throughArma adaptive prediction shows the necessity of controlling the qualitative behaviour of the adaptive filtering solutions.     

Theoretical studies on the metathesis processes, (Tp(PH3)MR(eta 2-H[bond]CH3)]-->[Tp(PH3)M(CH3)(eta 2-H[bond]R)] (M=Fe,Ru, and Os; R=H and CH3)

Theoretical calculations on the metathesis process, [Tp(PH3)MR(eta 2-H[bond]CH3)] --> [Tp(PH3)M(CH3)(eta 2-H[bond]R)] (M=Fe, Ru, and Os; R=H and CH3), have been systematically carried out to study their detailed reaction mechanisms. Other than the one-step mechanism via a four-center transition state and the two-step mechanism through an oxidative addition/reductive elimination pathway, a new one-step mechanism, with a transition state formed under oxidative addition, has been found. Based on the intrinsic reaction coordinate calculations, we found that the trajectories of the transferring hydrogen atom in the metathesis processes studied are similar to each other regardless of the nature of reaction mechanisms.     

Nitridation under ammonia of high surface area vanadium aerogels

Vanadium pentoxide gels have been obtained from decavanadic acid prepared by ion exchange on a resin from ammonium metavanadate solution. The progressive removal of water by solvent exchange in supercritical conditions led to the formation of high surface area V₂O₅, 1.6H₂O aerogels. Heat treatment under ammonia has been performed on these aerogels in the 450-900 °C temperature range. The oxide precursors and oxynitrides have been characterized by XRD, SEM, TGA, BET. Nitridation leads to divided oxynitride powders in which the fibrous structure of the aerogel is maintained. The use of both very low heating rates and high surface area aerogel precursors allows a higher rate and a lower threshold of nitridation than those reported in previous works. By adjusting the nitridation temperature, it has been possible to prepare oxynitrides with various nitrogen enrichment and vanadium valency states. Whatever the V(O,N) composition, the oxidation of the oxynitrides in air starts between 250 and 300 °C. This determines their potential use as chemical gas sensors at a maximum working temperature of 250 °C.     

A DFT study of SiH(4) activation by Cp(2)LnH

A theoretical study of SiH(4) activation by Cp(2)LnH complexes for the entire series of lanthanides has been carried out at the DFT-B3PW91 level of theory. The reaction paths corresponding to H/H exchange and silylation, formation of Cp(2)Ln(SiH(3)), have been computed. They both occur via a single-step sigma-bond metathesis mechanism. For the athermal H/H exchange reaction, the calculated activation barrier averages 1.8 kcal.mol(-)(1) relative to the precursor adduct Cp(2)LnH(eta(2)-SiH(4)) for all lanthanide elements. The silylation path is slightly exogenic (DeltaE approximately -6.5 kcal.mol(-1)) with an activation barrier averaging 5.2 kcal.mol(-1) relative to the precursor adduct where SiH(4) is bonded by two Si-H bonds. Both pathways are therefore thermally accessible. The H/H exchange path is calculated to be kinetically more favorable whereas the silylation reaction is thermodynamically preferred. The reactivity of this familly of lanthanide complexes with SiH(4) contrasts strongly with that obtained previously with CH(4). The considerably lower activation barrier for silylation relative to methylation is attributed to the ability of Si to become hypervalent.     

Efficient alkene hydrosilation with bis(8-quinolyl)phosphine (NPN) nickel catalysts. The dominant role of silyl-over hydrido-nickel catalytic intermediates

A cationic nickel complex of the bis(8-quinolyl)(3,5-di-tert-butylphenoxy)phosphine (NPN) ligand, [(NPN)NiCl]⁺, is a precursor to efficient catalysts for the hydrosilation of alkenes with a variety of hydrosilanes under mild conditions and low catalyst loadings. DFT studies reveal the presence of two coupled catalytic cycles based on [(NPN)NiH]⁺ and [(NPN)NiSiR₃]⁺ active species, with the latter being more efficient for producing the product. The preferred silyl-based catalysis is not due to a more facile insertion of alkene into the Ni–Si (vs. Ni–H) bond, but by consistent and efficient conversions of the hydride to the silyl complex.

A cationic nickel complex of the bis(8-quinolyl)(3,5-di-tert-butylphenoxy)phosphine (NPN) ligand, [(NPN)NiCl]⁺, is a precursor to efficient catalysts for the hydrosilation of alkenes with hydrosilanes under mild conditions and low catalyst loadings.     

Biscarbene-ruthenium complexes in catalysis: novel stereoselective synthesis of (1E,3E)-1,4-disubstituted-1,3-dienes via head-to-head coupling of terminal alkynes and addition of carboxylic acids

The reaction of a variety of alkynes RCtbd1;CH with a variety of carboxylic acids R(1)CO(2)H, in the presence of 5% of RuCl(COD)C(5)Me(5), selectively leads to the dienylesters (1E,3E)-RCH(1)=CH(2)-CH(3)=C(R)(O(2)CR(1)). The reaction also applies to amino acid and dicarboxylic acid derivatives. It is shown that the first step of the reaction consists of the head-to-head alkyne coupling and of the formation of the metallacyclic biscarbene-ruthenium complex isolated for R = Ph and catalyzing the formation of dienylester. D-labeled reactions show that the alkyne protons remain at the alkyne terminal carbon atoms and carboxylic acid protonates the C(1) carbon atom. QM/MM (ONIOM) calculations, supporting a mixed Fischer-Schrock-type biscarbene complex, show that protonation occurs preferentially at the carbene carbon C(1) adjacent to Ru, in the relative cis position with respect to the Ru-Cl bond, to give a mixed C(1)alkyl-C(4)carbene complex in which the C(4) carbene is conjugated with the noncoordinated C(2)=C(3) double bond. This 16-electron intermediate has a weak stabilizing alpha agostic C-H bond. This most stable isomer appears to have a C(4) center more accessible to the nucleophilic addition which accounts for the experimentally observed product.     

Practical synthesis of 1-deoxy-d-xylulose and 1-deoxy-d-xylulose 5-phosphate allowing deuterium labelling

An optimised gram scale synthesis allows the production of 1-deoxy-d-xylulose and 1-deoxy-d-xylulose 5-phosphate with possible deuterium labelling at C-5. Such substrates are required for investigations on the mevalonate-independent 2-C-methyl-d-erythritol 4-phosphate pathway for isoprenoid biosynthesis in bacteria and chloroplasts of phototrophic eukaryotes and for the biosynthesis of vitamins B₁ (thiamine diphosphate) and B₆ (pyridoxol phosphate) in bacteria.