In Statistical Machine Translation (SMT), inference needs to be performed over a high-complexity discrete distribution de ned by the intersection between a translation hypergraph and a target language model. This distribution is too complex to be represented exactly and one typically resorts to approximation techniques either to perform optimisation { the task of searching for the optimum translation { or sampling { the task of nding a subset of translations that is statistically representative of the goal distribution. Beam-search is an example of an approximate optimisation technique, where maximisation is performed over a heuristically pruned representation of the goal distribution. For inference tasks other than optimisation, rather than nding a single optimum, one is really interested in obtaining a set of probabilistic samples from the distribution. This is the case in training where one wishes to obtain unbiased estimates of expectations in order to t the parameters of a model. Samples are also necessary in consensus decoding where one chooses from a sample of likely translations the one that minimises a loss function. Due to the additional computational challenges posed by sampling, n-best lists, a by-product of optimisation, are typically used as a biased approximation to true probabilistic samples. A more direct procedure is to attempt to directly draw samples from the underlying distribution rather than rely on n-best list approximations. Markov Chain Monte Carlo (MCMC) methods, such as Gibbs sampling, o er a way to overcome the tractability issues in sampling, however their convergence properties are hard to assess. That is, it is di cult to know when, if ever, an MCMC sampler is producing samples that are compatible iii with the goal distribution. Rejection sampling, a Monte Carlo (MC) method, is more fundamental and natural, it o ers strong guarantees, such as unbiased samples, but is typically hard to design for distributions of the kind addressed in SMT, rendering an intractable method. A recent technique that stresses a uni ed view between the two types of inference tasks discussed here | optimisation and sampling | is the OS approach. OS can be seen as a cross between Adaptive Rejection Sampling (an MC method) and A optimisation. In this view the intractable goal distribution is upperbounded by a simpler (thus tractable) proxy distribution, which is then incrementally re ned to be closer to the goal until the maximum is found, or until the sampling performance exceeds a certain level. This thesis introduces an approach to exact optimisation and exact sampling in SMT by addressing the tractability issues associated with the intersection between the translation hypergraph and the language model. The two forms of inference are handled in a uni ed framework based on the OS approach. In short, an intractable goal distribution, over which one wishes to perform inference, is upperbounded by tractable proposal distributions. A proposal represents a relaxed version of the complete space of weighted translation derivations, where relaxation happens with respect to the incorporation of the language model. These proposals give an optimistic view on the true model and allow for easier and faster search using standard dynamic programming techniques. In the OS approach, such proposals are used to perform a form of adaptive rejection sampling. In rejection sampling, samples are drawn from a proposal distribution and accepted or rejected as a function of the mismatch between the proposal and the goal. The technique is adaptive in that rejected samples are used to motivate a re nement of the upperbound proposal that brings it closer to the goal, improving the rate of acceptance. Optimisation can be connected to an extreme form of sampling, thus the framework introduced here suits both exact optimisation and exact iv sampling. Exact optimisation means that the global maximum is found with a certi cate of optimality. Exact sampling means that unbiased samples are independently drawn from the goal distribution. We show that by using this approach exact inference is feasible using only a fraction of the time and space that would be required by a full intersection, without recourse to pruning techniques that only provide approximate solutions. We also show that the vast majority of the entries (n-grams) in a language model can be summarised by shorter and optimistic entries. This means that the computational complexity of our approach is less sensitive to the order of the language model distribution than a full intersection would be. Particularly in the case of sampling, we show that it is possible to draw exact samples compatible with distributions which incorporate a high-order language model component from proxy distributions that are much simpler. In this thesis, exact inference is performed in the context of both hierarchical and phrase-based models of translation, the latter characterising a problem that is NP-complete in nature.

ITC/USA 2011 Conference Proceedings / The Forty-Seventh Annual International Telemetering Conference and Technical Exhibition / October 24-27, 2011 / Bally's Las Vegas, Las Vegas, Nevada / A novel type-II hybrid automatic repeat request (HARQ) algorithm using adaptive modulations and bit-interleaved coded modulation (BICM) is presented. The algorithm uses different optimized puncturing patterns for different transmissions of the same data packet. The proposed approach exploits mapping diversity through BICM with iterative decoding. The modulation order is changed in each transmission to keep the number of symbols transmitted constant. We present new bit error rate and frame error rate analytical results for the proposed technique showing good agreement with simulation results. We compare the throughput performance of our proposed HARQ technique with a reference HARQ technique that uses different mapping arrangements but keeps the modulation order fixed. By using optimized puncturing patterns and adaptive modulations, our method provides significantly better throughput performance over the reference HARQ method in the whole signalto- noise ratio (SNR) range, and achieves a gain of 12 dB in the medium SNR region.

Please download dissertation to view abstract.

<p>Abstract</p><p>Title: The GI- method –bluff or the path to healthy life?</p><p>Number of pages: 41</p><p>Tutor: Lowe Hedman</p><p>Author: Annika Westlund</p><p>Course: Media and Communication Studies C</p><p>Period: Fall 2006</p><p>University: Division of Media and Communications Studies C</p><p>Purpose/Aim: The aim of this essay was to investigate how the site GI viktkoll describes glycemic index on their website. The intention was also to investigate how the media presented GI through articles and how they used doctors and dieticians to appear trustworthy. Another aim was to investigate what effect GI viktkoll could have on its readers.</p><p>Method/Material: I have chosen a qualitative method where I did a discourse analysis of the articles which were presented on the GI viktkolls website during a period of three weeks. This was my main method in the essay. I also did two interviews with educated professionals. The articles on the website were thereby the main material I used in the essay.</p><p>Main results: My result shows that GI viktkoll do have an underlying aim in wanting to influence its readers ina specific way. Therefore my result shows that it is important as a reader, to be aware of that GI viktkoll might not present a critical way of thinking and every aspect of the phenomenon. GI viktkoll also has influence on people because they have power to change peoples mind about the phenomenon GImethod in the society. GI viktkoll only presents the healthy way of living through the GI method, although there still there is a lot of disagreement from other directions such as doctors and dieticians about the actual effects of the GI method on healthy people.</p><p>Keywords: glycemic index, media culture, encoding/decoding, discourse analysis.</p>

<p>Abstract</p><p>Title: What TV-glasses do you wear? A study in how party-political people decode the TVshow Scooby Doo (Vilka tv-glasögon har du? En studie i hur partipolitiskt aktiva personer tolkar tv-serien Scooby Doo)</p><p>Number of pages: 47 (54 including enclosures)</p><p>Author: Christopher Landstedt</p><p>Tutor: Amelie Hössjer</p><p>Course: Media and Communication Studies C</p><p>Period: Autumn term 2007</p><p>University: Division of Media and Communication, Department of Information Science, Uppsala University.</p><p>Purpose/Aim: The aim of this essay is to make a study in how party-political people, 18-25 years old, both female and male, decode the messages in the TV-show Scooby Doo from 1969. Do they decode the show differently because of their political view, their gender or, and their social background? Is there a pattern in the decoding or is it based on a more individual level?</p><p>Material/Method: A qualitative method containing a total number of 16 individual interviews with young adults, 18-25 years old, half of them female, the other half male, were used. All of the participants are members of political youth parties/organizations, equally divided in left and right wing parties. Scooby Doo was chosen thanks to the lack of political meanings and messages in the show and its objective aura. The respondents got to see a preselected episode from the first season ever of Scooby Doo. After they finished watching the show, the interview took place. The interview contained questions on a deeper lever regardingthe episode. Stuart Hall’s all time classic encoding-decoding theory is used as the main theory with the support from other theories in the same field.</p><p>Main results: The degree of active reading is overall equal among the young adults that participated in the study. Differences can be found in the way they decode the sender’s messages and what values they put into the message. The leftwing respondents tended to decode the show in more oppositional way than the rightwing people who tended to read the messages dominant. There is an exception to every rule, also in this case. To sum it all up in one last sentence it should be said that some people’s personal values shine through, and aremore obvious than others.</p><p>Keywords: encoding-decoding, gender, television, interpretation, Scooby Doo, political view, leftwing and rightwing</p>

One channel model that characterises multipath fading effect of a wireless channel is called Flat Rayleigh Fading channel model. Given the properties of Flat Rayleigh Fading channel, an equation to find the capacity of a Flat Rayleigh fading channel with hard decision decoding is derived. The difference of power requirement to achieve the Additive White Gaussian Noise (AWGN) capacity over a Flat Rayleigh Fading channel fading is found to increase exponentially with Es /N0 . Upper and lower bounds of error performance of linear block codes over a Flat Rayleigh Fading channel are also studied. With the condition that the excess delay of a channel is known earlier, it is shown that a correlator with shorter length, according to excess delay of the channel, can be constructed for use in wireless channel response measurements. Therefore, a rule of construction of a shorter length correlator is defined, involving concatenation of parts of a Constant Amplitude Zero Auto-Correlation (CAZAC) sequence. Simulation of [136,68,24] Double Circulant Code with Dorsch List Decoding is also done in order to evaluate error performance of the channel coding scheme over one of the IEEE Wireless Metropolitan Area Network (WirelessMAN) channel models, the Stanford University Interim Channel Model No. 5 (SUI-5) channel. Performance of the channel cod- ing was severely degraded over the SUI-5 channel when it is compared to its performance over the AWGN channel. Indoor path losses within three multifloor office buildings were investigated at 433 MHz, 869 MHz and 1249 MHz. The work involved series of extensive received signal strength measurements within the buildings for all of the considered frequencies. Results have shown that indoor path loss is higher within a square footprint building than indoor path loss in a rectangular building. Parameters of Log-Distance Path Loss and Floor Attenuation Factor Path Loss models have been derived from the measurement data. In addition, a new indoor path loss prediction model was derived to cater for path loss pre- diction within multifloor buildings with indoor atriums. The model performs with better prediction accuracy when compared with Log-Distance Path Loss and Floor Attenuation Factor Path Loss models.

No description available.

We give an architecture of a storage system consisting of a storage medium made of unreliable memory elements and an error correction circuit made of a combination of noisy and noiseless logic gates that is capable of retaining the stored information with the lower probability of error than a storage system with a correction circuit made completely of noiseless logic gates. Our correction circuit is based on the iterative decoding of low-density parity check codes, and uses the positive effect of errors in logic gates to correct errors in memory elements. In the spirit of Marcus Tullius Cicero's Clavus clavo eicitur (one nail drives out another), the proposed storage system operates on the principle: error errore eicitur-one error drives out another. The randomness that is present in the logic gates makes these classes of decoders superior to their noiseless counterparts. Moreover, random perturbations do not require any additional computational resources as they are inherent to unreliable hardware itself. To utilize the benefits of logic gate failures, our correction circuit relies on two key novelties: a mixture of reliable and unreliable gates and decoder rewinding. We present a method based on absorbing Markov chains for the probability of error analysis, and explain how the randomness in the variable and check node update function helps a decoder to escape to local minima associated with trapping sets.

Le système d'identification automatique (ou système AIS pour automatic identification system) est un système qui permet aux navires et aux stations côtières de s'échanger certaines informations par radio VHF. Ces informations comprennent l'identifiant, le statut, la position, la direction et la vitesse de l'émetteur. L'objectif de cette thèse est de permettre la réception des messages AIS par un satellite en orbite basse sans modifier le matériel existant équipant les navires. Par l'intermédiaire du système AIS, il devient possible de connaitre la position de tous les navires à travers le monde. Plusieurs nouveaux services sont possibles, comme le contrôle maritime global ou, pour les armateurs, la connaissance constante de la position de leurs bateaux. La réception par satellite des signaux AIS est sujette à un niveau de bruit bien plus élevé que lors de la réception de ces signaux au niveau du sol. Ce niveau de bruit rend les méthodes classiques de réception de ces signaux difficilement utilisables. Une première contribution de cette thèse est le développement de nouveaux démodulateurs utilisant des méthodes de correction d'erreurs. Ceux-ci tirent parti de la présence d'un bloc de contrôle de redondance cyclique (CRC) dans les messages ainsi que de certaines informations connues sur la structure des messages et des données. Des adaptations du récepteur proposé ont également été étudiées afin d'intégrer la poursuite de la phase des signaux reçus et de prendre en compte les collisions des messages envoyés simultanément par plusieurs navires. La dernière partie de cette thèse est consacrée à l'étude des méthodes de localisation des navires ne diffusant pas leur position dans leurs messages AIS. Cette localisation tire parti des paramètres des messages reçus tels que le délai de propagation et le décalage en fréquence de la porteuse dû à l'effet Doppler, et d'un modèle de déplacement des navires. / The automatic identification system (AIS) is a system allowing ships and coast stations to exchange some information by VHF radio. This information includes the identifier, status, location, direction and speed of the emitter. The aim of this thesis is to allow the reception of AIS messages by low Earth orbit satellites without modifying the existing ship equipments. With this system, it becomes possible to know the position of all ships over the Earth. As a consequence, several new services become available, such as global traffic monitoring or determining boat location (for ship-owners). Satellite reception of AIS signals is subjected to a higher noise level when compared to ground level reception. This noise makes classical demodulation and decoding methods unusable. A first contribution of this thesis is to develop new demodulators using error correction methods. These demodulators take advantage of the presence of a cyclic redundancy check (CRC) block in the messages as well as known information about the structure of messages and data. Generalizations of the proposed receiver have also been studied in order to take into account the phase noise of the received signals and the possible collision of messages sent simultaneously by several vessels. The last part of this thesis is devoted to the study of localization methods for ships that do not transmit their location in AIS messages. This localization takes advantage of information contained in the received messages such as the propagation delay and the carrier frequency shift due to the Doppler effect, and a ship movement model.

Research question: 1. How have famous people in images contributed to the spread of #BlackLivesMatter’s message? 2. How can the content in the image be tied to the BlackLivesMatter movement.   Aim: The aim of this study is to research how famous people have appeared in images and in a what way they have contributed to spread of #BlackLivesMatter’s message   Method and material: A case study will be performed on the collected material from  two specific case-events that are affiliated to the movement. The material will be analyzed with a semiotic method with guidelines by Hansen and Machin (2013). All material is collected online and consists pictures that can be connected to BlackLivesMatter on a denotative or connotative level.   Main results: This study's result showed that different types of performances and appearances with famous people contributed to spreading BlackLivesMatter’s message. The contribution was visualized through performances, posing at events and music videos, and use of different objects such as clothing, people, hands, and/or posing that created connotative association which could be connected to the BlackLivesMatter movement.