Gotowa bibliografia na temat „Transfert asynchrone”

Following prolonged exposure to audiovisual asynchrony, an observer’s point of subjective simultaneity (PSS) shifts in the direction of the leading modality. It has been debated whether other sensory pairings, such as vision and touch, lead to a similar temporal recalibration, and if so, whether the internal timing mechanism underlying lag visuotactile adaptation is centralised or distributed. To address these questions, we adapted observers to vision- and tactile-leading visuotactile asynchrony on either their left or right hand side in different blocks. In one test condition, participants performed a simultaneity judgment on the adapted side (unilateral) and in another they performed a simultaneity judgment on the non-adapted side (contralateral). In a third condition, participants adapted concurrently to equal and opposite asynchronies on each side and were tested randomly on either hand (bilateral opposed). Results from the first two conditions show that observers recalibrate to visuotactile asynchronies, and that the recalibration transfers to the non-adapted side. These findings suggest a centralised recalibration mechanism not linked to the adapted side and predict no recalibration for the bilateral opposed condition, assuming the adapted effects were equal on each side. This was confirmed in the group of participants that adapted to vision- and tactile-leading asynchrony on the right and left hand side, respectively. However, the other group (vision-leading on the left and tactile-leading on the right) did show a recalibration effect, suggesting a distributed mechanism. We discuss these findings in terms of a hybrid model that assumes the co-existence of a centralised and distributed timing mechanism.

A total of 78 day 10 horse embryos were transferred non-surgically to recipient mares that had ovulated 9, 7, 6, 5, 4, 3, 2 or 1 day after (negative asynchrony), on the same day (synchronous), or 2 or 4 days before (positive asynchrony) the donor (n=6 or 8 mares per group). Pregnancy rates between 100% (6/6) and 63% (5/8) were seen in recipient mares that were between +2 and −6 days asynchronous. Embryo survival to the heartbeat stage declined in recipients that were −7 days asynchronous and no embryos survived in recipients that were −9 days asynchronous. Irrespective of uterine asynchrony, cessation of embryo mobility and fixation at the base of a uterine horn occurred when the conceptus was ∼17 days old. Conceptus growth and development was slowed when embryos were placed in negatively asynchronous uteri. At the greatest degree of negative asynchrony at which embryos survived to the heartbeat stage, i.e. −7 and −6 days, development of the embryo proper and allantois was retarded. Luteostasis was achieved in recipient mares when day 10 embryos were transferred to recipient mares at any stage of asynchrony between −9 and +2 days with respect to the donor. These results indicate that in the horse, there is a wide window for establishment of pregnancy following embryo transfer to asynchronous recipients. Although progesterone priming of the uterus to a stage equivalent to that of the transferred embryo does not appear to be a prerequisite for embryo survival, it does nonetheless influence embryonic development.

The aim of this study was to determine the effect of maternal–embryonic asynchrony in the reproductive tract (oviduct and uterus) on subsequent embryo development in cattle. Fifty Day 1invitro-produced zygotes were transferred endoscopically into the oviduct ipsilateral to the corpus luteum of heifers (n=40) that were either synchronous with the embryos (Day 1 after ovulation) or asynchronous and ahead of the embryo (Day 3 after ovulation). A subset of heifers was killed in a commercial abattoir 3, 6 or 14 days after embryo transfer. Location within the reproductive tract, developmental stage and the quality of embryos were recorded. Transfer of embryos to an advanced (asynchronous) oviduct resulted, on Day 4, in fewer embryos at the expected location (oviduct), and a greater number of degenerated and retarded embryos with a lower total cell number than for embryos in the synchronous group. Similarly, on Day 7, asynchrony led to a greater number of degenerated and retarded embryos compared with the synchronous group. Total embryo cell number was similar among groups. Although Day 15 conceptuses were longer following asynchronous transfer, only 50% of the asynchronous heifers yielded conceptuses, compared with 100% in the synchronous group. In conclusion, asynchrony between the developing embryo and the reproductive tract has a negative effect on embryo development.

Abstract STUDY QUESTION Does the reproductive potential of embryos change when blastocyst development takes longer than the traditionally accepted 5 days when accounting for aneuploidy and endometrial-embryo asynchrony? SUMMARY ANSWER Aneuploidy increases with increasing duration of blastulation, but if blastocyst morphologic quality and endometrial-embryo asynchrony are controlled for, euploid Day 7 embryos have similar sustained implantation as compared to Days 5 and 6 euploid blastocysts. WHAT IS KNOWN ALREADY The relative contributions of diminished embryo quality versus endometrial and embryo asynchrony to poor outcomes associated with embryos cultured past Day 6 are not clear. Asynchrony can be eliminated by embryo vitrification with transfer in a subsequent month after retrieval. STUDY DESIGN, SIZE, DURATION Retrospective cohort study of patients from a single center attempting conception through ICSI and utilizing preimplantation genetic testing for aneuploidy screening (PGT-A) from January 2017 to September 2018. Cycles were excluded if they utilized surgical sperm or preimplantation genetic testing for monogenetic/single gene defects. ICSI cycle outcomes from 2586 patients were evaluated for ploidy status of embryos. PARTICIPANTS/MATERIALS, SETTING, METHODS Only patients undergoing single, euploid frozen embryo transfer were included when analyzing cycle outcomes by day of blastocyst expansion of the transferred embryo (n = 2130). Ploidy rates by the day upon which an embryo was considered to be usable (denoted, ‘usable blastulation day’) were determined so as to assess the contribution of aneuploidy to slow embryo development. Outcomes of euploid frozen single embryo transfers (SET) of Day 7 embryos were evaluated to assess the reproductive potential associated with embryos that were slowly developing for reasons other than aneuploidy. Analyses were adjusted by maternal age and blastocyst morphology. MAIN RESULTS AND THE ROLE OF CHANCE Overall, 67.7% (n = 3508) of usable Day 5 blastocysts were euploid, 52.1% (n = 5560) of usable Day 6 blastocysts were euploid and 43.1% (n = 229) of usable Day 7 embryos were euploid (Day 5 versus Day 6: odds ratio (OR) 0.7 (95% CI, 0.64–0.76), P < 0.001; Day 5 versus Day 7: OR 0.56 (95% CI, 0.46–0.69), P < 0.001; Day 6 versus Day 7: OR 0.81 (95% CI, 0.67–0.99), P = 0.036). Stratified by Society for Assisted Reproductive Technology maternal age groups, a reduction in the prevalence of euploidy by increasing time to embryo blastulation was still seen. The sustained implantation rate (SIR) was similar after euploid SET of Days 5 and 6 embryos (overall, 68.9% (95% CI, 66.0–71.6) and 66.8% (95% CI, 63.8–69.7), respectively; P = 0.81). SIR after euploid Day 7 SET appeared slightly lower than that of Days 5 and 6 embryos (52.6% (95% CI, 35.8–69.0); (Day 5 versus Day 7: OR, 0.67 (95% CI, 0.32–1.41), P = 0.29; Day 6 versus Day 7: OR 0.58 (95% CI, 0.28–1.2), P = 0.14)) but did not achieve statistical significance. LIMITATIONS, REASONS FOR CAUTION The primary limitation is the low number of Day 7 blastocyst transfers that limits statistical power. Additionally, the retrospective nature of this study may prevent full elucidation of potential biases with respect to culture, morphologic assessment and selection of Day 7 embryos for transfer. WIDER IMPLICATIONS OF THE FINDINGS Routine culture through Day 7 may successfully increase the pool of transferrable embryos for patients who would otherwise have no usable embryos if culture terminated on Day 6. This is particularly true for older patients (i.e. greater than 35 years of age), whose embryos take longer to blastulate and, therefore, are more susceptible to cycle cancelation. Additionally, as evidenced by an adequate overall SIR of 52.6% after euploid SET of Day 7 blastocysts, embryos developing to a usable blastocyst on Day 7 are likely within the ‘window of blastulation.’ STUDY FUNDING/COMPETING INTEREST(S) None.

Background: Comparative features of embryos developed under in vitro and in vivo conditions are particularly important in designing embryo transfer procedures that fulfil embryo-recipient synchronization requirements. Objective: To determine the degree of asynchrony in rabbit embryo development between cultured and in vivo embryos. Methods: A total of 55 non-lactating multiparous female rabbits were used. Embryos were classified as 16-cells or early morulae at 48 hours post-coitum (hpc). Embryos were cultured during 30 or 32 h and embryo development was compared with in vivo embryos of 72 hpc. In vitro and in vivo embryos at 72 hpc were classified as early or compacted morulae. Bayesian statistics was used. Difference between in vivo and in vitro embryos and the actual probability of the difference between the in vivo and in vitro embryo higher than zero (P) was estimated. Results: The percentage of compacted morulae was higher in in vivo embryos than in in vitro embryos with +6 h of asynchrony (73.5 and 32.8%, P=1.00). But the percentage of compacted morulae was similar with +8 h asynchrony. Conclusions: In vitro embryos delay their development by + 8 hours compared to in vivo embryos.

Mechanical ventilation is an essential life-support treatment for patients who cannot breathe independently. Patient–ventilator asynchrony (PVA) occurs when ventilatory support does not match the needs of the patient and is associated with a series of adverse clinical outcomes. Deep learning methods have shown a strong discriminative ability for PVA detection, but they require a large number of annotated data for model training, which hampers their application to this task. We developed a transfer learning architecture based on pretrained convolutional neural networks (CNN) and used it for PVA recognition based on small datasets. The one-dimensional signal was converted to a two-dimensional image, and features were extracted by the CNN using pretrained weights for classification. A partial dropping cross-validation technique was developed to evaluate model performance on small datasets. When using large datasets, the performance of the proposed method was similar to that of non-transfer learning methods. However, when the amount of data was reduced to 1%, the accuracy of transfer learning was approximately 90%, whereas the accuracy of the non-transfer learning was less than 80%. The findings suggest that the proposed transfer learning method can obtain satisfactory accuracies for PVA detection when using small datasets. Such a method can promote the application of deep learning to detect more types of PVA under various ventilation modes.

Substantial evidence now suggests that a positive diversity–stability relationship exists. Yet few studies examine the facets of biodiversity that contribute to this relationship, and empirical research is predominantly conducted on grassland communities under controlled conditions. We investigate the roles of species richness, environmental condition (vegetation cover), asynchrony, and weighted population stability in driving community stability across multiple taxa. We used data from a Long-term Ecological Research project to investigate temporal stability of annual plants, beetles, reptiles, and rodents in Nizzanim Coastal Sand Dune Nature Reserve in Israel. All four taxa had a strong positive relationship between asynchrony and community stability. Only rodents showed a positive richness–stability relationship. Perennial plant cover had a significant relationship with community stability for three taxa, but the direction of the correlation varied. Asynchrony had a stronger relationship with perennial plant cover than it did with richness for both plants and beetles. We suggest that community stability is driven by asynchrony for flora as well as fauna. Stability appears to be determined by species’ interactions and their responses to the environment, and not always by diversity. This has important consequences for understanding the effects of environmental degradation on ecosystem stability and productivity, which have destabilizing consequences beyond biodiversity loss.

The timing of estrus synchrony between donor and recipient does is an important consideration in an embryo transfer program. Experiments were conducted to determine the optimal time of estrus synchrony between donor and recipient dairy goats used in a commercial nuclear transfer (NT) program. Donor and recipient synchronization was achieved by implanting either a 3-mg norgestomet ear implant (Crestar®, Intervet Int. B.V., Boxmeer, Holland) or a 300-mg progesterone vaginal implant (CIDR-G®, Pharmacia and Upjohn Ltd. Co., Auckland, NZ) on Day 0. A single 5mg intramuscular injection of prostaglandin (Lutalyse®, Pharmacia and Upjohn, Kalamazoo, MI, USA) was administered on Day 7. Recipients received a single 200–500IU intramuscular injection of PMSG (Calbiochem, LaJolla, CA, USA) on Day 13. Alternatively, starting on Day 12 donors received twice daily intramuscular injection (64mg/day) of FSH (Folltropin®, Vetrepharm, Ontario, Canada) over four consecutive days. On Day 15 the implants were removed from both donors and recipients and the animals were mated several times daily to vasectomized bucks over two consecutive days. In Experiment 1, estrus synchrony or asynchrony was achieved by removing the implant from recipients at the same time or 12h later than donors, respectively. In Experiment 2, only estrus asynchrony was utilized and was achieved by removing the implant from recipients either 12 or 18h later than donors. In vivo-ovulated MII oocytes surgically recovered from superovulated donors on Day 17 were enucleated and reconstructed with transfected caprine fetal or adult skin cells or transgenic adult skin cells. Couplets were simultaneously fused, activated, and then cultured in SOF/BSA for 48h at 38°C. Two-to-eight-cell NT embryos at 48h post-fusion and activation were surgically transferred to the oviducts of surrogate recipients with similar implant types and PMSG doses. Pregnancies were determined by ultrasonography starting at approximately Day 28 post-fusion and activation and then monitored weekly. In Experiment 1, there were significantly more pregnant asynchronous recipients compared with synchronous recipients (6 of 24 v. 12 of 124 does, respectively). While there were no significant differences, more offspring were produced per embryo transferred to asynchronous recipients compared with synchronous recipients (5 of 135 v. 11 of 690 offspring per embryo transferred, respectively). In Experiment 2, while not significant, there were more pregnant +12-h asynchronous recipients compared with +18-h asynchronous recipients (16 of 72 v. 5 of 36 does, respectively). Again, while there were no significant differences, more offspring were produced per embryo transferred to +12h compared with +18h asynchronous recipients (11 of 424 v. 3 of 224 offspring per embryo transferred, respectively). These results suggest that asynchrony of estrus between recipients and donors is more beneficial in a commercial caprine NT program, and that +12h may be a more optimal period of asynchrony for recipient does receiving NT embryos. Table 1 Summary of recipient estrus synchronization

Equine pregnancy is characterised by an unusually long pre-implantation period (40 days) during which the conceptus is entirely dependent on uterine secretions for nutrient provision; although glucose is an important nutrient during development post-blastocyst formation, little is known about its transport into the early horse conceptus. Equine embryos are also known to tolerate an unusually large degree of uterine asynchrony following embryo transfer (ET). However, negative asynchrony (recipient behind the donor) of more than 5 days markedly retards conceptus growth and development, and thereby offers a unique tool for studying the effect of the uterine environment on early development. In a preliminary study, we detected abundant mRNA expression for the facilitative glucose transporters (SLC2As) 1–3, 5, 8 and 10 and sodium-glucose co-transporter (SLC5A)11 in Day 14 to 28 equine conceptus membranes. In the current study, we evaluated the effect of uterine asynchrony on trophectodermal glucose transporter expression. Day 8 horse embryos were transferred to recipient mares that ovulated on the same day (synchronous; n = 10) or 5 days after (asynchronous; n = 10) the donor mare. The conceptuses were collected 6 or 11 days after ET (Day 14 or 19 of embryo development: n = 5 per group). Trophectodermal mRNA expression for glucose transporters was evaluated by RT-qPCR, and the effects of asynchronous ET and stage of pregnancy were analysed by two-way ANOVA followed by independent-samples t-tests. Gene expression for SLC2A3 and 8 was stable over time and treatment. Expression of SLC2A1 and SLC5A11 decreased between Days 14 and 19 in synchronous pregnancies only (P < 0.05). SLC2A2 expression increased markedly on Day 19 in synchronous (P < 0.01) but not asynchronous pregnancies (P < 0.05). SLC2A5 expression was lower in the asynchronous group on Day 14, but increased beyond expression levels in synchronous pregnancies by Day 19 (P < 0.05). In summary, expression of SLC2A1 and 3, the major placental glucose transporters, was not affected by asynchronous ET. The marked up-regulation of SLC2A2 expression between Days 14 and 19 of synchronous but not asynchronous pregnancy suggests a stage-specific function, whereas the increase in SLC2A5 at Day 19 after asynchronous ET could be a compensatory response to growth retardation. This study was funded by EpiHealthNet (Project number 317146).

La faible duree de vie des produits et la multitude des variantes proposees aux clients obligent les entreprises a disposer de systemes de production flexibles et modulaires. Dans ce travail de recherche, les systemes d'assemblage, particulierement representatifs puisqu'ils constituent souvent l'une des dernieres etapes du processus de production ou se fait la personnalisation, sont utilises comme support. On se focalise plus precisement sur l'aspect pilotage du flux dans le cas d'une ligne flexible d'assemblage a transfert asynchrone. En effet, outre les difficultes inherentes a la conception d'un systeme flexible d'assemblage, la gestion du flux en cours d'exploitation necessite, pour chaque nouvelle production, la determination optimale des parametres de pilotage afin de minimiser les delais, les encours, les reglages, etc l'objectif de ce travail est de proposer une architecture d'integration d'outils informatiques permettant d'optimiser les parametres de pilotage d'une ligne en exploitation. La simulation de flux est le premier outil utilise. Son temps de reponse, tres nettement inferieur a celui du systeme reel, permet de tester un grand nombre de combinaisons des parametres. Mais si le nombre de combinaisons est tres important, la simulation exhaustive de chacune d'elles s'avere longue et fastidieuse. On propose donc de coupler la simulation a un autre outil : un algorithme d'optimisation de type heuristique. Les experimentations realisees avec plusieurs algorithmes et en particulier avec un algorithme genetique, montrent l'efficacite de ce couplage. Pour satisfaire simultanement plusieurs objectifs, l'algorithme est ensuite complete par une methode de selection multicritere.

Cette thèse concerne la conception et la validation de plusieurs mécanismes de contrôle de flux pour le service ABR dans les réseaux ATM. L'algorithme max-min représente une des méthodes les plus utilisées pour offrir le critère de l'équité entre les utilisateurs. La première contribution de cette thèse consiste à étendre l'algorithme max-min à une version pondérée baptisée WMM. Selon cette extension, nous avons donc proposé deux algorithmes distribués : l'algorithme mico et l'algorithme mclf. Ces deux derniers améliorent considérablement les performances de l'algorithme erica a savoir l'équité entre les utilisateurs. En présence des sources ABR sporadiques, L'utilisation conjointe des comportements de source, des algorithmes de commutateur et des comportements use-it-or-lose-it définis par l'ATM forum, ne permet pas de contrôler efficacement la sporadicité du trafic en provenance de ces sources. Elle entraîne, par conséquent, une surestimation de l'allocation de la bande passante aux sources ABR. Dans cette perspective, nous avons conçu l'algorithme iaaa qui calcule l'ARC en fonction de l'état actuel du trafic afin de réduire l'effet provoqué par la sporadicité. Cette thèse a été achevée par la mise en oeuvre de deux algorithmes rca et rca#+. Ils offrent la possibilité d'acheminer un trafic de type vidéo en utilisant le service ABR. Pour ce faire, une équation entre le paramètre de quantification q et le volume de l'information à transmettre a été ajustée. Cette étude s'est basée sur une allocation équitable des sources ABR, sur un contrôle de débit du codeur vidéo, ainsi que sur une diminution du taux de réduction de l'image. Ceci a permis un meilleur délai de transfert tout en gardant une qualité d'image satisfaisante. Ces différents algorithmes ont fait l'objet d'une étude de simulation destinée à valider leurs performances.

Le travail réalisé dans le cadre de cette thèse porte sur la transformation automatique de programmes synchrones en vue de leur implémentation asynchrone sur les architectures hétérogènes distribuées. Dans un contexte applicatif nous nous sommes plus particulièrement intéressés aux systèmes de traitements temps-réel de l'image. La démarche que nous avons suivie se base sur une phase préliminaire de conception et validation des systèmes temps-réel sous l'hypothèse de synchronisme. En pratique, nous avons utilisé un outil existant : le langage signal. Nous avons développé un ensemble de processus de transformation permettant d'automatiser l'implémentation de ces systèmes sous la forme de processus séquentiels communicants. Les caractéristiques de ce travail sont de garantir a priori la sûreté des programmes en transformant des systèmes prouves, de répartir intégralement les échanges de données de manière à ne pas recourir à des mécanismes centralises, de conserver à l'exécution un modèle dicté par la disponibilité des données. Les systèmes ainsi transformés sont constitués d'un ensemble fini de fonctions de calcul dotées d'interfaces de communications ; elles constituent les tâches du système. Ces tâches ne communiquent entre-elles que par l'intermédiaire de leurs entrées et sorties. L'émission et la réception de données sont conditionnées par des commandes gardées implémentées dans le corps des interfaces

Dans un réseau quel qu'il soit, le contrôle de trafic est une composante des plus importantes. Son existence est liée à la possibilité d'avoir des situations de congestion qui peuvent provoquer des effondrements complets des performances du réseau. Devant la complexité croissante des réseaux due notamment à la diversité des services supportés, les mécanismes de contrôle de trafic classiques montrent leurs limites. Une des idées maîtresses de cette thèse consiste à montrer la nécessité de s'orienter vers une approche de contrôle adaptatif qui permet de tenir compte des aspects évolutifs du réseau. Dans cette étude, nous nous sommes particulièrement intéressés au problème de contrôle de congestion dans les réseaux atm. Ces travaux considèrent les systèmes multi-agents comme un outil capable de répondre aux besoins d'adaptabilité et tenant compte des aspects de distribution inhérents aux réseaux de télécommunications. De par les capacités d'adaptation et d'interaction qui les caractérisent, les systèmes multi-agents constituent des architectures particulièrement adaptées à supporter des systèmes de contrôle adaptatif. Un modèle qui permet de mettre en oeuvre des mécanismes de contrôle flexibles et efficaces a été défini. Il est structuré en couches et constitue de deux niveaux : un niveau contrôle et un niveau meta-contrôle. Le premier niveau assure les fonctions de contrôle de base. Le second niveau, implémenté par des agents, apporte au niveau contrôle la flexibilité nécessaire. Pour mettre en oeuvre la couche contrôle, nous avons proposé le mécanisme trac (threshold algorithm for control) qui utilise des seuils de congestion afin de declencher différentes actions de contrôle. Le mécanisme d-trac (dynamic trac) résulte de l'intégration d'agents au mécanisme trac. Il exploite les capacités d'interaction et d'observation des agents pour effectuer un ajustement dynamique des seuils de congestion et une exécution flexible des actions de contrôle. Un environnement de simulation de réseau atm intégrant des agents a été développé en utilisant des environnements de développement agent (le langage ptitloo et la plate-forme dima).

Dans cette thèse on propose une chaîne de communication performante dans un contexte où les contraintes sur la QoS et sur l'efficacité spectrale visée ne peuvent être satisfaites par les systèmes actuels. On se place dans un contexte de communications satellitaires où l'on transmet des paquets de petite taille. Classiquement, les seuls systèmes envisageables pour la plage d'efficacité spectrale [0. 5,0. 8] font cohabiter des codes et des modems incompatibles en terme de performances. Le système proposé dans cette thèse est basé sur la définition de la constellation 3PSK possédant des caractéristiques situées entre la BPSK et la QPSK mais une capacité de démodulations supérieure. Pour permettre un bon couplage code-modem on propose d’associer un modem burst ternaire à des codes ternaires TCP et LDPC performants. On montre enfin, après optimisation du système global code-modem, que sur la plage d'efficacité spectrale cible les performances outrepassent celles de son équivalent binaire
The aim of this thesis consist in proposing a powerful communication system for satellite transmissions where the constraints put on the QoS and the spectral efficiency cannot be handled by existing systems. We work in a context of satellite communications where small packets are transmitted. Typically, the combinations code-modem which are proposed for spectral efficiencies in [0. 5,0. 8] are unbalanced in term of performance. The system we propose in this thesis is based on the 3PSK constellation which possesses a better demodulation capacity than BPSK or QPSK. In order to improve the cohabitation code-modem, we propose to associate a ternary burst modem with powerful ternary codes (TPC & LDPC). Finally, we show, after optimizations of the global system, that in the range of spectral efficiencies [0. 5,0. 8], our systems exhibit better performances than their binary counterparts

Les mécanismes de coopération sont de plus en plus important dans les communications sans fil et réseaux permettant d'accroître les performances du système et ce en diminuant la consommation d'énergie, en augmentant la capacité du système et en rendant les taux de perte de paquets plus petits. L'idée de coopération peut être retracée par une étude en théorie de l'information sur le canal à relais dans le réseau cellulaire. D'un point de vue système, puisque un mobile (MS) a des limites de taille (antenne unique), de puissance, de coût et de matériel, il devient irréalisable d'utiliser la technologie MIMO pour un MS. Les utilisateurs mobiles avec antennes uniques peuvent encore profiter de la diversité spatiale par le biais de transmissions coopératives codées en espace temps. L'objectif de cette thèse est d'introduire et de discuter de diverses stratégies de coopération dans les communications sans fil. Dans la première partie, nous présentons une analyse de performances de bout en bout d'une communication coopérative asynchrone et à saut multiple, utilisant des relais regénératifs évoluant dans un canal de Rayleigh en évanouissement plat. Le schéma de conception est basé sur un précodage sur les trames qui seront encodés par paires. Ce précodage est basé sur l'ajout d'un préfixe cyclique, contenant une séquence d'apprentissage. Nous développons le taux d'erreur sur les bits et le taux d'erreur sur les bites de bout en bout relativement à une modulation BPSK, pour des canaux à gains égaux et à gains différents. Nous présentons aussi les performances en taux d'erreur sur les trames et le taux d'erreur sur les trames de bout en bout. Finalement, nous présentons une comparaison entre trois configurations que nous avons introduit, différenciées par le degré de coopération. La deuxième partie contient deux chapitres. Dans le premier chapitre, nous considérons un système dans lequel un relais choisit de coopérer que si le canal source-relais est d'une qualité acceptable et nous évaluons l'utilité du relayage lorsque la source agit aveuglément et ignore la décision des relais s'ils peuvent coopérer ou non. Dans notre étude, nous considérons de relais regénératifs pour lesquels les décisions de coopération sont fondées sur un rapport signal à bruit (SNR) seuil et d'examiner l'éventuel impact d'une détection et transmission d'erreurs par le relais. Nous dérivons le taux d'erreur sur les bits (BER) de bout en bout pour une modulation BPSK et nous examinons deux stratégies de répartition de puissance entre la source et le relais afin de minimiser le BER de bout en bout au niveau de destination pour un ordre de SNR élevé. Dans le second chapitre, nous consédérons un schéma dans lequel un relais choisit de coopérer seulement lorsque le canal source-destination est d'une qualité inaccepable. Dans notre étude, nous considérons une relais regénératif pour lequel la décision de coopération est basée sur un seuil du rapport signal à bruit et nous considérons les effets d'une éventuelle détection et transmission d'erreur par le relais. Nous développns l'expression du taux d'erreur sur les bits pour une modulation BPSK et nous examinons une stratégie optimale permettant de minimiser ce BER de bout en bout pour un orde de SNR élevé. Dans la troisième partie, nous considérons un canal à accès multiple MAC où deux utilisateurs communiquent avec une destination commune. Chaque utilisateur agit mutuellement comme un relais pour l'autre ainsi qu'il tient à transmettre ses propres informations, par opposition à une consacration totale au relayage. Nous tenons à évaluer l'utilité du relais d'un point de vue somme des taux du système (sum rate) plutôt que d'un point de vue de l'utilisateur bénéficiant de la coopération qui se fait habituellement. Nous faisons cela en permettant un compromis entre le relayage et la transmission de données via une d'allocation des ressources. Concrètement, nous proposons un régime de transmission coopérative permettant à chaque utilisateur d'allouer une certaine fraction de puissance pour sa propre transmission de données tandis que le reste est consacré au relayage. Le protocole sous-jacent est basé sur une modification du protocole Non-Orthogonal Amplify and Forward (NAF). Nous développons les expressions de capacité pour notre modèle et nous présentons une allocation de puissance optimisant une forme approche du sum rate, pour les cas centralisé et distribué. Dans le scénario distribué, une connaissance hybride de l'état du canal, partie statistique et partie instantanée, est exploitée. L'algorithme d'allocation de puissance dans le cas centralisé indique que, même dans une coopération mutuelle comme la notre, quelle que soit la réalisation du canal, la coopération mutuelle n'est jamais réelle, c'est-à-dire l'un des utilisateurs allouera toujours une puissance nulle pour relayer les données de l'autre, et donc agit égoïstement. Mais dans le cadre distribué, nos résultats indiquent que la somme des taux est maximisée lorsque les deux mobiles agissent égoïstement
Cooperative mechanisms are becoming increasingly important in wireless communications and networks to substantially enhance system performance with respect to much less power consumption, higher system capacity and smaller packet loss rate. The idea of cooperation can be traced back to the information theory investigation on relay channel in cellular network. From the system point of view, since Mobile Station (MS) has limitations in single antenna, power, cost and hardware, it is infeasible to use MIMO technology in MS. Mobile users with single antennas can still take advantage of spatial diversity through cooperative space-time encoded transmission. The objective of this thesis is to introduce and discuss various cooperative strategies in wireless communications. In the first part, we present an end-to-end performance analysis of two-hop asynchronous cooperative diversity with regenerative relays over Rayleigh block-flat-fading channels, in which a precoding frame-based scheme with packet-wise encoding is used. This precoding is based on the addition of a cyclic prefix which is implemented as a training sequence. We derive, for equal and unequal sub-channel gains, the bit-error rate and the end-to-end bit-error rate expressions for binary phase-shift keying. We also present the performance of the frame-error rate and the end-to-end frame-error rate. Finally, comparisons between three system configurations, differing by the amount of cooperation, are presented. The second part contains two chapters. In the first chapter, we consider a scheme in which a relay chooses to cooperate only if its source-relay channel is of an acceptable quality and we evaluate the usefulness of relaying when the source acts blindly and ignores the decision of the relays whether they may cooperate or not. In our study, we consider the regenerative relays in which the decisions to cooperate are based on a signal-to-noise ratio (SNR) threshold and consider the impact of the possible erroneously detected and transmitted data at the relays. We derive the end-to-end bit-error rate (BER) for binary phase-shift keying modulation and look at two power allocation strategies between the source and the relays in order to minimize the end-to-end BER at the destination for high SNR. In the second chapter, we consider a scheme in which the relay chooses to cooperate only if the source-destination channel is of an unacceptable quality. In our study, we consider a regenerative relay in which the decision to cooperate is based on a signal-to-noise ratio (SNR) threshold and consider the effect of the possible erroneously detected and transmitted data at the relay. We derive an expression for the end-to-end bit-error rate (BER) of binary phase-shift keying (BPSK) modulation and look at the optimal strategy to minimize this end-to-end BER at the destination for high SNR. In the third part, we consider a multiple access MAC fading channel with two users communicating with a common destination, where each user mutually acts as a relay for the other one as well as wishes to transmit his own information as opposed to having dedicated relays. We wish to evaluate the usefulness of relaying from the point of view of the system's throughput (sum rate) rather than from the sole point of view of the user benefiting from the cooperation as is typically done. We do this by allowing a trade-off between relaying and fresh data transmission through a resource allocation framework. Specifically, we propose cooperative transmission scheme allowing each user to allocate a certain amount of power for his own transmitted data while the rest is devoted to relaying. The underlying protocol is based on a modification of the so-called non-orthogonal amplify and forward (NAF) protocol. We develop capacity expressions for our scheme and derive the rate-optimum power allocation, in closed form for centralized and distributed frameworks. In the distributed scenario, partially statistical and partially instantaneous channel information is exploited. The centralized power allocation algorithm indicates that even in a mutual cooperation setting like ours, on any given realization of the channel, cooperation is never truly mutual, i. E. One of the users will always allocate zero power to relaying the data of the other one, and thus act selfishly. But in distributed framework, our results indicate that the sum rate is maximized when both mobiles act selfishly

Cette étude présente la problématique de l'interconnexion de cartes de commutation dans les commutateurs ATM. L'objectif est de proposer une solution permettant d'augmenter la capacité globale de commutation de ces commutateurs. L'analyse des architectures des principaux prototypes et commutateurs industriels montre trois niveaux d'organisation comprenant les modules de commutation, les cartes de commutation et l'interconnexion de celles-ci. L'étude montre que la principale limite à la capacité globale de commutation de ces architectures provient des structures d'interconnexion des cartes. La comparaison des performances des structures d'interconnexion par réseau de clos, chaînage, en anneau et en bus fait ressortir les avantages de ce dernier en termes de débit global d'interconnexion, de capacité d'expansion, et de modularité. Il présente d'autre part des possibilités intéressantes pour la duplication nécessaire de cellules dans les services de diffusion d'information. L'importance des débits d'interconnexion oriente naturellement les solutions vers l'utilisation de fibres optiques pour la réalisation de ces bus. L'architecture d'une carte de commutation intégrant une interface d'interconnexion par bus est proposée. Cette carte utilise un module de commutation BSS adapté à la réalisation de commutateurs de forte capacité de commutation. Les performances présentées en fin d'étude montrent l'homogénéité d'une telle architecture et le potentiel que l'on peut en attendre

This contribution explores the workings of linguistic (im)politeness in plot advancement in the context of audiovisual translation with a focus on English-Korean film subtitling, which represents a transfer between considerably remote linguacultures that rely on very different pragmalinguistic means of expressing interpersonal meanings. Challenges of interlingual subtitling are associated with cultural asynchrony both between the source and target cultures and between the verbal component recreated in a different language and the original semiotic system. The discussion is based on examples from relatively recent mainstream fiction films that demonstrate some of the solutions deployed by screen translators in order to capture and re- or co-create the interpersonal effects of the original that have direct bearing on the general plot development.

To accommodate the needs of large-scale distributed systems, scalable data storage and management strategies are required, allowing applications to efficiently cope with continuously growing, highly distributed data. This chapter addresses the key issues of data handling in grid environments focusing on storing, accessing, managing and processing data. We start by providing the background for the data storage issue in grid environments. We outline the main challenges addressed by distributed storage systems: high availability which translates into high resilience and consistency, corruption handling regarding arbitrary faults, fault tolerance, asynchrony, fairness, access control and transparency. The core part of the chapter presents how existing solutions cope with these high requirements. The most important research results are organized along several themes: grid data storage, distributed file systems, data transfer and retrieval and data management. Important characteristics such as performance, efficient use of resources, fault tolerance, security, and others are strongly determined by the adopted system architectures and the technologies behind them. For each topic, we shortly present previous work, describe the most recent achievements, highlight their advantages and limitations, and indicate future research trends in distributed data storage and management.

Spatial frequency detection (Dt) and discrimination (Ds) thresholds were measured by a 2 × 2 AFC procedure. The Dt/Ds ratio was shown (1) to be independent of stimulus duration (from 10 to 1000 ms), (2) but to depend on the stimulus onset asynchrony (SOA) between the test stimulus and a suprathreshold compound mask (composed of the two frequencies to be discriminated) with a maximum ratio for SOAs of ~25 ms (backward masking). These findings are true independent of the objective spatial frequency difference between the stimuli to be discriminated. Nevertheless, (3) when the SOA is fixed at 25 ms and the contrast of the mask is increased, the Dt/Ds ratio shows a correlative increase only if it is significantly greater than 1 (i.e., perfect discrimination) in the absence of the mask. The three sets of results can be accounted for in terms of a serial processing of Dt and Ds where (i) the time constant of the Ds-stage is shorter than or equal to the time constant of the Dt-stage; (ii) the transfer of information from one stage to another requires ~25 ms, and (iii) the transducer of the DS-stage is nonlinear with respect to the transducer of the Dt-stage.