Using Monte Carlo simulations and analytic methods, we examine the dynamics of inter-species competition using the Penna bit-string model. We begin with a study of the steady state with a single species, then proceed to the dynamics of competition between two species. When the species are not evenly matched in fitness, a simple differential equation provides a satisfactory model of the behavior of the system. However, when the species are equally fit, we show that a model, originally proposed to describe population genetics [Fisher,Wright], is required. When mutations are allowed between the competing species, the dynamics becomes more interesting. The mutation rate becomes a parameter that dictates the steady state behavior. If the two species are not equally fit, the value of the mutation rate determines whether the longer-lived or faster reproducing species is favored. With two species that are equally fit, the steady state varies with mutation rate from a single peaked to a double peaked distribution. This behavior is shown to be well described by an extension to the Fisher-Wright model mentioned above. Finally, we describe the preliminary results of a few new lines of investigation, and suggest ideas for further study of the dynamics of this model. / Ph. D.

Small intestinal ileus affects up to half of all horses undergoing small intestinal surgery, leading to prolonged gastrointestinal (GI) transit time which can be life-threatening. Various prokinetic medications have been associated with varying side effects, questionable efficacy, and increased cost. Gum chewing as a form of sham feeding is used as a safe, effective, well-tolerated, and inexpensive way to ameliorate ileus following GI surgery in humans. Bit chewing for horses, an analogous activity, has been shown to significantly decrease GI total transit time (TTT); however, a direct effect of bit chewing on gastric emptying time (GET), small intestinal transit time (SITT), and total orocecal transit time (OCTT) has not been investigated. Our objective was to determine whether bit chewing increased small intestinal motility and decreased GE, SITT, and OCTT in clinically normal horses. Gastrointestinal motility was compared in horses that were bit chewing compared to control conditions (no bit chewing) in a prospective crossover design study using acetaminophen as a marker for GET and video endoscopy (ALICAM) capsules to determine GET, SITT, and OCTT. Bit chewing was well tolerated by all horses with no side effects noted. Bit chewing led to a shorter GET, SITT, and significantly shorter OCTT when compared to the control group (P = 0.015). Median times for bit chewing conditions were as follows: GE 2.86 hr, SITT 3.65 hr, and OCTT 6.15 hr whereas the median times for control conditions were as follows: GE: 5 hr, SITT 4.4 hr, and OCTT 9.92 hr. In summary, bit chewing proves to be a potential tool to hasten the motility of the oral GIT. It is safe, inexpensive, and potentially effective prokinetic treatment to horses suffering from postoperative ileus and further investigation is warranted. / Master of Science / Ileus, or a temporary lack of intestinal motility, is a common life-threatening problem in horses, especially following abdominal surgery. Current treatments have questionable efficacy and high cost. In human patients suffering from ileus, sham feeding with gum chewing that mimics normal food consumption with no food ingested has shown promising results in improving clinical signs and restarting gastrointestinal motility. Bit chewing, an analogous activity in horses, also decreases gastrointestinal (GI) total transit time (TTT). However, ileus in horses typically affects the small intestine, which is a part of the GI tract that has not been investigated in regards to bit chewing. Our objective was to determine whether bit chewing shortens gastric emptying time (GET), small intestinal transit time (SITT), and total orocecal transit time (OCTT) in clinically normal horses. Gastrointestinal motility was compared in horses that were bit chewing compared to control (no bit chewing) conditions in a prospective crossover design study using acetaminophen serum samples as a marker for GET and video endoscopy (ALICAM) capsules to determine GET, SITT, and OCTT. There were no adverse effects to bit chewing and OCTT was significantly shortened in horses when bit chewing compared to the control group. In summary, bit chewing is a tool to hasten gastrointestinal motility, specifically small intestinal motility, and it may be a safe, inexpensive, and effective treatment to improve small intestinal motility in horses suffering from ileus and further investigation is warranted.

Echo state network (ESN) is a specific neural network structure composed of high dimensional nonlinear dynamics and learned readout weights. This thesis considers applying ESN for symbol detection in multiple-input, multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) systems. A new ESN structure, namely, windowed echo state networks (WESN) is introduced to further improve the symbol detection performance. Theoretical analysis justifies WESN has an enhanced short-term memory (STM) compared with the standard ESN such that WESN can offer better computing ability. Additionally, the bandwidth spent as the training set is the same as the demodulation reference signals defined in 3GPP LTE/LTE-Advanced systems for the ESN/WESN based symbol detection. Meanwhile, a unified training framework is developed for both comb and scattered pilot patterns. Complexity analysis demonstrates the advantages of ESN/WESN based symbol detector compared to conventional symbol detectors such as linear minimum mean square error (LMMSE) and sphere decoder when the system is employed with a large number of OFDM sub-carriers. Numerical evaluations show that ESN/WESN has an improvement of symbol detection performance as opposed to conventional methods in both low SNR regime and power amplifier (PA) nonlinear regime. Finally, it demonstrates that WESN can generate a better symbol detection result over ESN. / Artificial neural networks (ANN) are widely used in recognition tasks such as recommendation systems, robotics path planning, self-driving, video tracking, image classifications, etc. To further explore the applications of ANN, this thesis considers using a specific ANN, echo state network (ESN) for a wireless communications task: MIMO-OFDM symbol detection. Furthermore, it proposed an enhanced version of the standard ESN, namely, windowed echo state network (WESN). Theoretical analyses on the short term memory (STM) of ESN and WESN show that the later one has a longer STM. Besides, the training set size of this ESN/WESN based method is chosen the same as the pilot symbols used in conventional communications systems. The algorithm complexity analysis demonstrates the ESN/WESN based method performs with lower complexity compared with conventional methods, such as linear mean square error (LMMSE) and sphere decoding. Comprehensive simulations examine how the symbol detection performance can be improved by using ESN and its variant WESN when the transmission link is non-ideal.

Made available in DSpace on 2016-04-04T18:31:21Z (GMT). No. of bitstreams: 1 Marcelo Aparecido Morales.pdf: 2213720 bytes, checksum: 6858d1301ee70451e07970bdb465607e (MD5) Previous issue date: 2008-02-15 / The IEEE 802.11 wireless local area network presents a MAC anomaly when stations with different bit rates are connected in the same Access Point. Stations with high SNR have worst performance than stations with low SNR. This condition is worse in PWLANs (Public Wireless Local Area Network) with users connecting with different bit rates. This paper proposes a policy that uses the SNR and the condition of propagation to control users Contention Window. With this policy it is possible to control the Bit Rate, which is not possible in the 802.11 networks. / Redes 802.11 apresentam uma anomalia quando existem esta??es com diferentes taxas conectadas a um mesmo ponto de acesso. Esta??es com boa SNR (Signal-to-Noise Rate) obt?m um desempenho pior que esta??es com pior SNR. Esta condi??o ? esperada de forma agravada em redes p?blicas com v?rios usu?rios se conectando a diferentes taxas. Neste trabalho ? proposta uma pol?tica atrav?s da varia??o da janela de conten??o, levando em considera??o a SNR e a condi??o de propaga??o do local. Com a pol?tica a ser apresentada ? poss?vel controlar a taxa de transmiss?o em fun??o da SNR, criando uma possibilidade n?o oferecida pela rede 802.11.

Recently, there has been a renewed interest in code based cryptosystems. Amongst the reasons for this interest is that they have shown to be resistant to quantum at- tacks, making them candidates for post-quantum cryptosystems. In fact, the National Institute of Standards and Technology is currently considering candidates for secure communication in the post-quantum era. Three of the proposals are code based cryp- tosystems. Other reasons for this renewed interest include e cient encryption and decryption. In this dissertation, new code based cryptosystems (symmetric key and public key) are presented that use high rate codes and have small key sizes. Hence they overcome the drawbacks of code based cryptosystems (low information rate and very large key size). The techniques used in designing these cryptosystems include random bit/block deletions, random bit insertions, random interleaving, and random bit ipping. An advantage of the proposed cryptosystems over other code based cryp- tosystems is that the code can be/is not secret. These cryptosystems are among the rst with this advantage. Having a public code eliminates the need for permutation and scrambling matrices. The absence of permutation and scrambling matrices results in a signi cant reduction in the key size. In fact, it is shown that with simple random bit ipping and interleaving the key size is comparable to well known symmetric key cryptosystems in use today such as Advanced Encryption Standard (AES). The security of the new cryptosystems are analysed. It is shown that they are immune against previously proposed attacks for code based cryptosystems. This is because scrambling or permutation matrices are not used and the random bit ipping is beyond the error correcting capability of the code. It is also shown that having a public code still provides a good level of security. This is proved in two ways, by nding the probability of an adversary being able to break the cryptosystem and showing that this probability is extremely small, and showing that the cryptosystem has indistinguishability against a chosen plaintext attack (i.e. is IND-CPA secure). IND-CPA security is among the primary necessities for a cryptosystem to be practical. This means that a ciphertext reveals no information about the corresponding plaintext other than its length. It is also shown that having a public code results in smaller key sizes. / Graduate

Real-time optimization of drilling parameters during drilling operations aims to optimize weight on bit, bit rotation speed for obtaining maximum drilling rate as well as minimizing the drilling cost. The process is considered to be formation specific. A statistical method such as multiple linear regression technique has been used for the drilling optimization. An extensive literature survey on drilling optimization was conducted for this research study. A model is developed for this purpose using actual field data collected through modern well monitoring and data recording systems, which predicts the rate of drilling penetration as a function of available parameters. The rate of penetration general equation is optimized for effective functions at each data point. In order to optimize the parameters in the field, a computer network is required to be developed. The computer network will keep the piped data directly from the data source, and continuously be collecting the new data to be fed. A database present at the central computer will be continuously calculating the developed model parameters by means of multiple regression technique and inform the team at the field. The field engineer will transmit the current drilling parameters back to the central computer, and the headquarters will determine the new model parameters and optimum drilling parameters by including the recently received information. Therefore, there will be a real-time-optimization process. It is considered that this technique is going to be widely used in future drilling activities since it could reduce drilling costs and minimize probability of encountering problems due to working with optimized parameters. It has been found that drilling rate of penetration could be modelled in real-time environment as a function of independent drilling variables such as weight on bit, rotation speed of the string, drilling fluid weight, and formation characteristics. The ability to have the drilling rate of penetration with respect to depth characteristically with certain parameters for specific formations on real-time basis could bring new insights to the nature of drilling optimization studies. Any significant departure of the actual rate of penetration from the predicted rate of penetration trend could have important indications which could be detected beforehand in real-time. The study has also achieved one of its objectives, giving the optimized independent drilling parameters found following statistical synthesis.

La compréhension de la dynamique du transport électronique dans des conducteurs balistiques et cohérents est indispensable à la réalisation d’expériences d’optique électronique ou de calcul quantique à partir de « flying qu-bits ». La première étape est de pouvoir injecter en régime d’effet Hall quantique un électron dans la mer de Fermi sans excitation supplémentaire : un léviton, dont les propriétés remarquables ont été expérimentalement démontrées sans champ magnétique [1], ainsi que contrôler sa trajectoire à l’aide d’interrupteurs quantiques. Dans ce travail de thèse, nous avons réalisé l’étape préliminaire qui valide la possibilité de créer des lévitons en régime d’effet Hall Quantique : cela consiste à démontrer la validité de la théorie de bruit photo-assisté dans ce régime, en utilisant une excitation sinusoïdale et monochromatique, plus simple et plus contrôlée, que celle conduisant aux lévitons. En outre, nous avons étudié lors de la thèse le phénomène physiquement relié de l’interrupteur quantique élémentaire, qui est l’ouverture et la fermeture très soudaine d’un canal de conduction élémentaire. Ce phénomène, qui pose la question fondamentale « Que se passe-t-il lorsque que la mer de Fermi est spatialement coupée en deux ? », génère un bruit intrinsèque de charge [2] que nous avons mis en évidence et donne une mesure théorique de l’entropie d’intrication quantique.[1] J. Dubois, T. Jullien, F. Portier, P. Roche, A. Cavanna, Y. Jin, W. Wegscheider, P. Roulleau, and D. C. Glattli. minimal-excitation states for electron quantum optics using levitons. Nature, 502(7473), October 2013.[2] Israel Klich and Leonid Levitov, Phys. Rev. Lett. 102, 100502 (2009) / The study of dynamic electronic transport in ballistic coherent conductor is required for the implementation of electron quantum optics experiments or the quantum computation by using “flying qu-bits”. The first step is to be able to inject in the quantum Hall effect a single electron without any additional excitations in the conductor: a Leviton, whose remarkable properties have been experimentally observed without magnetic field [1], and to control its trajectory thanks to a quantum switch. During this thesis, we confirmed the possibility to implement a leviton in the quantum Hall effect, by demonstrating the validity of the photo-assisted shot-noise theory in this regime: we use a sine excitation which is simpler to implement than a lorentzian excitation required for Levitons. We also studied a new effect described by the photo-assisted theory: a quantum switch, which is the sudden closing and opening of an elementary channel of conduction. This generates an intrinsic charge noise [2] that we have evidenced and enable us to answer the general question: “What are the effects of a spatial separation of the Fermi sea?”. Furthermore, this charge noise provides a theoretical measurement of the entanglement entropy.[1] J. Dubois, T. Jullien, F. Portier, P. Roche, A. Cavanna, Y. Jin, W. Wegscheider, P. Roulleau, and D. C. Glattli. minimal-excitation states for electron quantum optics using levitons. Nature, 502(7473), October 2013.[2] Israel Klich and Leonid Levitov, Phys. Rev. Lett. 102, 100502 (2009)

This report documents the design and construction of a computer-controlled engraver (CNC engraver). Whereas most common machines used for this purpose are dimensioned towards milling, this project aims to create a dedicated engraver which could be of reduced size and complexity in comparison to standard CNC routers. This is partly because engraving imparts no need for a spindle motor, and that the machine only needs to withstand forces related to engraving, which are often significantly smaller than those present when milling. The proposed construction is designed towards use on aluminium and retains about the size of a laptop computer. Additionally, it should allow for engraving on material of almost any size, which is not a common feature in CNC machines. The design uses Arduino and open-source software to generate and execute G-code and is built with off-the-shelf components as well as both 3D-printed and machined components. Key conclusions are that it is indeed possible to make such a machine and that it could fulfil some useful purposes, even if the prototype, as built, has some issues with structural rigidity and therefore the final tolerances of the engraving. These issues primarily stem from the assembly supporting the Z-Axis and engraving bit, which lacks the stiffness needed for desirable results. Taking the fairly low complexity of the build and the price tag of under 2000 SEK in consideration, the prospects are deemed as promising. / Denna rapport beskriver konstruktionen och byggandet av en datorkontrollerad (CNC) gravyrmaskin. Emedan de gängse maskinerna för detta ändamål är en vanlig fräs eller dylikt, så syftar detta projekt att bygga en dedikerad gravyrmaskin, som då jämförelsevis kan vara av reducerad storlek och komplexitet. Maskinen använder Arduino och öppen källkod för att generera och exekvera G-kod, och byggs med inköpta standardkomponenter, såväl som både 3D-printade och egentillverkade komponenter. De främsta slutsatserna är att det mycket riktigt går att bygga en sådan maskin, och det finns användningsområden för en sådan, även om prototypen som den slutligen blev har problem med styvhet, och därför även med noggrannheten i gravyrerna.

<p dir="ltr"> The widespread adoption of Convolutional Neural Networks (CNNs) in real-world applications, particularly on resource-constrained devices, is hindered by their computational complexity and memory requirements. This research investigates the application of pruning and quantization techniques to optimize CNNs for arrhythmia classification using the MIT-BIH Arrhythmia Database. By combining magnitude-based pruning, regularization-based pruning, filter map-based pruning, and quantization at different bit-widths (4-bit, 8-bit, 2-bit, and 1-bit), the study aims to develop a more compact and efficient CNN model while maintaining high accuracy. The experimental results demonstrate that these techniques effectively reduce model size, improve inference speed, and maintain accuracy, adapting them for use on devices with limited resources. The findings highlight the potential of these optimization techniques for real-time applications in mobile health monitoring and edge computing, paving the way for broader adoption of deep learning in resource-limited environments.</p>

Drilling is one of the important aspects in the oil and gas industry due to the high demand for energy worldwide. Drilling time is considered as the major part of the operations time where the penetration rate (ROP) remains as the main factor for reducing the time. Maximizing ROP to lower the drilling cost is the main aim of operators. However, high ROP if not controlled may impact on the well geometry in terms of wellbore instability, cavities, and hole diameter restrictions. Accordingly, more time is needed for the other operations that follow such as: pool out of hole (POOH), casing running, and cementing. Bit wear is considered as the essential issue that influences in direct way on the bit performance and reduce ROP. Predicting the abrasive bit wear is required to estimate the right time when to POOH to prevent any costly job to fish any junk out to the surface. The two-common types of bits are considered in the research, rock bits (roller cone bits) and Polycrystalline Diamond Compact bits (PDC). This study focuses more on PDC bits because about 60% of the total footage drilled in wells worldwide were drilled by PDC bits and this is expected to reach 80% in 2020. The contribution of this research is to help reducing the drilling cost by developing new tools not to estimating the time when to POOH to surface but also to measure the wear and enhance the accuracy of prediction the bit efficiency. The work is broken down into four main stages or models to achieve the objective: The first stage; estimating of the rock abrasiveness and calculate the dynamic dulling rate of the rock bit while drilling. The second stage; estimating the PDC abrasive cutters wear by driving a new model to determine the mechanical specific energy (MSE), torque, and depth of cut (DOC) as a function of effective blades (EB). The accuracy of the predicted wear achieves 88% compared to the actual dull grading as an average for bits used in five wells. The third stage; modifying the previous MSE tool to develop a more accurate approach; effective mechanical specific energy (EMSE), to predict the PDC bit efficiency in both the inner and outer cone to match the standard bit dulling. The fourth stage; predicting ROP while PDC drilling in hole by accounting three parts of the process: rock drillability, hole cleaning, and cutters wear. The results achieve an enhancement of about 40% as compared to the available previous models. Consequently, the developed models in this study provide a novelty on understanding in more details the bit rock interface process and gain an idea of the relationship between the drilling parameters to enhance the bit performance and avoid damaging the bit. This is basically about optimisation the controllable factors such as: weight on bit (WOB), rotary speed (RPM), and flow rate. The result is the reduction in time losses and the operations cost. To ensure reliability and consistency of the proposed models, they were validated with several vertical oil wells drilled in Libya. The results from the validation of the models are consistent with the real field data. The research concludes that the developed models are reliable and applicable tool for both: to assist decision-makers to know when to pull the bit out to surface, and also to estimate the bit performance and wear.