Publikationer från Uppsala Universitet
Not a member yet
127603 research outputs found
Sort by
AI-stödd kognitiv radioteknik för förbättrad satellitkommunikation
No full textThis thesis explores the application of AI-based cognitive radio technology to improve spectral efficiency and interference mitigation in satellite communication systems. Satellite communication is necessary for global communication, but faces challenges such as spectrum congestion and signal interference due to the rapidly increasing number of devices using the spectrum. Cognitive radio technology in collaboration with AI offers promising solutions for adaptive and intelligent spectrum management. Several methods for signal detection and classification were developed. Two methods were created to detect user signals by analyzing their energy in the frequency domain, successfully identifying signal presence across multiple frequency bands.A reinforcement learning method combined with long-short-term memory (LSTM) was also investigated to enable real-time adjustments of transmission parameters. Due to evaluation challenges and practical limitations, this approach was set aside. Further efforts focused on distinguishing between normal and anomalous signals. An autoencoder based model was first developed using MATLAB but due to computational constraints, the development was transitioned to Python. The resulting model achieved a classification accuracy of approximately 75% for normal signals and 66% for anomalous signals.Finally, a supervised neural network (NN) was implemented to differentiate between clean and interference-affected signals, achieving a classification accuracy of around 96% for normal signals and almost 100% for signals with interference as the anomaly. To further evaluate the performance of the supervised NN, a classical method using matched filters was also implemented. While the classical method failed to detect interference when its frequency completely overlapped with the user signal, the supervised neural network successfully identified the interference in such cases.The methods presented in this work contribute to the field of AI-drived cognitive radio technology by presenting methods capable of detecting spectrum occupancy, optimizing signal throughput, and identifying interfering signals with varying levels of accuracy. These contributions aim to support more resilient, interference-aware satellite communication systems. I dagens samhälle används trådlös kommunikation mer och mer. Det kan handla om internet på avlägsna platser, nödsignaler ute till havs eller väderövervakning från rymden. Dessa kräver alla kommunikation med satelliter. Antal användare ökar för varje dag vilket skapar en konkurrens om radiospektrumet som i sin tur leder till störningar och sämre prestanda. I detta examensarbete undersöks hur artificiell intelligens (AI) kan användas för att göra satellitkommunikation smartare och mer effektiv. Genom att kombinera AI med kognitiv radioteknik så kan man lära system att anpassa sig själva. I projektet togs flera metoder fram för att upptäcka och klassificera olika signaler och avikelser. Två av metoderna analyserar signalers energi i olika frekvensområden för att avgöra vilka delar av spektrumet som används, och vilka delar som är lediga. En annan metod använde AI för att automatiskt ändra signalens parametrar när förutsättningarna i omgivningen förändras. Den metoden visade sig dock vara svår att utvärdera och använda i verkligheten. Ytterligare metoder togs fram för att kunna upptäcka signaler som avviker från det normala. Sådana avvikelser kan tyda på olika typer av störningar, fel eller attacker. En första modell gjordes i MATLAB men flyttades sedan över till Python för att kunna tränas fortare. Denna modell kunde känna igen normala signaler med ungefär 75% träffsäkerhet och signaler med störningar med ungefär 66%. Slutligen utvecklades en mer avancerad modell tänkt att kunna användas för olika typer av avvikelser. Denna modell kunde känna igen både normala signaler och signaler med störningar med en precision på över 95%. Denna slutliga modell jämfördes sedan med en klassisk metod med matchade filter. Där blev det tydligt att den nya slutliga modellen bättre klarar av att identifiera störare som ligger väldigt nära användarens signal i spektrumet. Detta arbete visar på att det finns stor potential i att utveckla kongitiv radio med hjälp av AI för att kunna förbättra satellitkommunikation. Detta är ett steg mot framtidens mer robusta och intelligenta kommunikationssystem
Clinical paper Emergency medical services, treatment of cardiac arrest patients and cardiac arrest registries in Europe - Update on systems
No full textIntroduction: Incidence and survival rates following cardiac arrest vary significantly across Europe. While several studies have sought to address the knowledge gap in the epidemiology of out-of-hospital cardiac arrest (OHCA), they have not successfully identified the reasons behind these disparities. This study aims to provide an updated overview of European Emergency Medical Systems (EMS). Methods: A questionnaire consisting of 35 main questions was used. The survey encompassed topics related to ambulance and dispatch characteristics, on-scene cardiac arrest management, as well as the availability and scope of datasets in cardiac arrest registries. Results: Survey responses were received from 27 European countries. While there were differences in the proportion of staff with advanced life support skills between countries, these staff were almost invariably dispatched in the event of a cardiac arrest call. First responder systems were available in only 17 countries. There were huge differences in ambulance control models, with the number of dispatch centres ranging from 0.4 to 42.2 per million population. Nine countries reported having out-of-hospital registries of prehospital cardiac arrest with national coverage while only three countries had registries of in-hospital cardiac arrest with full coverage. Conclusions: There are differences in EMS structures and the management of OHCA across Europe. Understanding these discrepancies is essential for improving OHCA outcomes and fostering greater uniformity in emergency response protocols throughout the region. Although there has been an increase in the population covered by a cardiac arrest registry, there is still a need to expand registry coverage, especially for registries of in-hospital cardiac arrest
Costs and Consequences of Chronic Kidney Disease in People with Diabetes in Portugal : A Modelling Study
No full textIntroduction: Chronic kidney disease is the fastest-growing chronic disease in terms of prevalence and one of the biggest causes of global mortality according to the Global Burden of Disease Collaboration. This study aimed to project the natural disease progression of this disease in people with diabetes, and to quantify the costs and consequences in the Portuguese context. This was achieved by developing an analytical model reflecting the epidemiology of chronic kidney disease and integrating the various stages of disease progression. Methods: A population-based cohort Markov model was used, to follow an adult cohort of people with diabetes and chronic kidney disease as they progressed through different risk categories, in annual cycles, over a period of 50 years. The model considered the natural progression of chronic kidney disease through 18 risk categories based on the KDIGO classification system, as well as the probability of patients receiving renal replacement therapy, including dialysis and kidney transplantation, and the probability of death. Each stage is associated with an annual cost and a disability weight, so the model allowed survival, years lived with disability and lifetime costs to be estimated for the entire population with chronic kidney disease and for patients in different risk categories. Results: Over the cohort<acute accent>s lifetime, the model estimated, for the total population with chronic kidney disease and diabetes, an average survival of 8.62 years, with 0.59 years lived with disability, and an average cost of <euro>24 613. These figures correspond to a loss of more than 410 000 years lived with disability and a total lifetime cost of 17.0 billion euros. The progression of this disease was associated with lower survival, more years lived with disability and higher costs. Conclusion: The results of this study characterize the natural progression of chronic kidney disease in people with diabetes mellitus type 2, as well as the associated costs and consequences in the national context. Since diabetes mellitus type 2 is a risk factor for chronic kidney disease, it is expected that the real impact will be greater than estimated in the coming decades. Analysis by risk level shows that progression of the disease is associated with worse outcomes
From Permian To Post-Modern: Changing Climate And Hab-Itability In A Warming World
No full textThe climate on planet Earth is a forever-changing, dynamic system. From the tiniest microorganismsto the largest planetary-scale processes, countless factors drive changes across timescales ranging frommilliseconds to millions of years. While many of these drivers act slowly and subtly over geologicalepochs, others can trigger rapid and far-reaching transformations. In modern times, climate change islargely attributed to anthropogenic emissions from fossil fuel use, with rising atmospheric CO2 levelsas a central driver. The socioeconomic pathways SSP370 and SSP585 project that CO2 concentrationscould reach 710 to 900 ppm respectively in Post-Modern Earth (PM) by the year 2080 — roughly threetimes higher than in the Pre-Industrial (PI) era of the 1850s — without involvement from large-scaleplanetary processes. Similar atmospheric conditions have existed in Earth’s distant past. One suchepoch is the Pre–End Permian (PEP), which occurred approximately 250 million years ago. Duringthis period, atmospheric CO2 levels are likewise estimated to have reached between 710 to 900 ppm.However, the planetary conditions were drastically different, as nearly all landmasses were clustered intoa single supercontinent known as Pangaea, the days were 2 hours shorter than today and Earth received2% less solar radiation.In this paper we present a comparative study between three distinct periods — Pre-End Permian (PEP),Pre-Industrial (PI), and Post-Modern (PM) — to explore the influence of planetary-scale changes andanthropogenic CO2 emissions on Earth’s climate and inhabitability. By focusing on the magnitude ofchange rather than timescale, we isolate the effects of these climate drivers using K¨oppen-Geiger cli-mate classification and Wet-Bulb Temperature as key indicators. Computed climate simulations usingROCKE-3D for each scenario are used to evaluate shifts in land area fractions of climate zones andinhabitable regions. Our findings show that increasing CO2 concentrations from PI to PM scenarioscorrelate with global warming and expanding the areas of uninhabitability from the equator toward thepoles. A transition toward more arid climates with the reduction in ice-covered and tundra zones is alsoshown. When comparing PEP and PM using 710 and 900 ppm CO2 our findings show that planetary-scale changes give rise to larger and more uneven changes of the K¨oppen-Geiger climate classificationsthan anthropogenic CO2 emissions. The extent of arid- and cold climate decreases while the extent oftropical, temperate and polar regions increases from PEP to PM for both 710 and 900 ppm. These find-ings suggest that the overall magnitude of climate change is more strongly influenced by planetary-scaledrivers than by anthropogenic emissions alone. To further evaluate spatial and temporal patterns in thesimulated data, we applied a modified Mann-Kendall trend test. This non-parametric method, adaptedfor gridded climate data, was used to detect significant monotonic trends in both spatial distributionand mean climate variables. The results confirm that planetary-scale changes generally cause more pro-nounced shifts in both climate trends and mean values compared to those driven by rising atmosphericCO2 levels. Finally, we simulate Earth’s climate with transmission spectra to see if any changes in cli-mate can be seen through the eye of spectroscopy. No discernible spectral differences are found betweenthe epochs, suggesting that changes in Earths climate remain undetectable with this method
"Har man tid över får man gärna lära sig sitt modersmål" : Legitimeringar, positioneringar och språkideologier i debatter om modersmålsundervisning sedan de allmänna valen 2022
No full textModersmålsundervisningen har länge fungerat som markör för inställningen till invandring och mångfald. Efter riksdagsvalet 2022 ingick Sverigedemokraterna, ett parti med tydligt motstånd i frågan, ett samarbete med regeringen, inom vilket undervisningens negativa konsekvenser skulle utredas. Denna förändring gör det intressant att undersöka debatten i frågan sedan dess. Tidigare undersökningar som har behandlat debatter om undervisningen har främst varit inriktade på hur den förs i tidningar, särskilt insändare. Fokus i denna uppsats är på argumenterande texter (tidningsartiklar och riksdagsprotokoll) med politiker som avsändare. Syftet är att beskriva de diskurser om modersmålsundervisning som framkommer och på så sätt öka förståelsen för det politiska samtalet om flerspråkighet. Därför analyseras, med en kritisk diskursanalytisk ansats, legitimeringar och delegitimeringar, positioneringar samt språkideologier. Generellt legitimeras undervisningen främst med hänvisning till annat som gynnas (till exempel utveckling i svenska och språk generellt), medan delegitimeringar ofta görs med hänvisning till bristande forskning och att samhället inte bör lägga resurser på undervisningen. Genom sina positioneringar framställer debattörerna till exempel sig själva och varandra som partirepresentanter. Språkideologiskt framträder ståndpunkter på en skala mellan en generell en- och flerspråkighetssyn. Resultaten diskuteras bland annat utifrån modersmålsundervisningens plats i utbildningspolitiken (vilket kan påverka vilka aspekter som aktualiseras i debatterna) och dess fortsatta symboliska kraft
Proteomic Profiling of Inflammation Following Traumatic Brain Injury and Cerebral Insults : Decoding Molecular Signatures for Therapeutic Insights
No full textTraumatic brain injury (TBI) is a complex and diverse condition influenced by various factors, all of which contribute to a wide range of outcomes for the patients. The initial injury renders the brain susceptible to secondary molecular and pathophysiological events, with neuroinflammation considered a crucial secondary mechanism. In this thesis, we investigated the inflammatory response following TBI and drew comparisons with central nervous system (CNS) inflammatory disorders. The aim was to identify distinct inflammatory responses specific to TBI pathology, ultimately contributing to precision medicine and advancing the development of individualized therapies. We used cerebral microdialysis (CMD) and proximity extension assay (PEA) to analyze protein expressions in patients with acute severe TBI. Additionally, we conducted mass-spectrometry (MS) analysis on cerebrospinal fluid (CSF) collected from patients with TBI and individuals with various types of encephalitis. Paper I - The PEA technique revealed the presence of inflammatory proteins in the brain interstitium during the initial days following severe TBI. Diverse temporal patterns were observed, with some proteins peaking early, others showing middle or late peaks, and some exhibiting biphasic or steady patterns. Notably, while these temporal patterns varied among proteins, unique and consistent patterns were discerned for the same proteins across different patients. Paper II - 21 proteins were selected for a customized panel to explore their temporal dynamics in patients suffering from acute TBI. This structured approach brought the method closer to clinical application and identified proteins of interest for monitoring in this patient group. Paper III - Patients had their CSF analyzed with the use of MS. The resulting protein findings from patients with acute severe TBI were then compared to findings from patients with non-infectious inflammatory CNS disorders, infectious inflammatory CNS disorders and controls. We found distinct proteomic profiles in the CSF of TBI patients compared with the healthy controls and different disease controls, indicating a specific interplay between inflammatory factors, metabolic response, and cell integrity. In relation to primarily infectious or inflammatory disorders, unique inflammatory pathways seem to be engaged and could potentially serve as future treatment targets. Paper IV - CSF from patients with acute severe TBI and COVID-19 patients with neurological symptoms were analysed using MS. Neurofilament light chain (NFL) levels were measured separately. Distinct inflammatory profiles in severe TBI and COVID-19 patients were identified. Elevated NFL levels in both groups indicated white matter injury. The discovery of novel biomarkers specific to each pathology, provide valuable insights into the inflammatory mechanisms and may serve as targets for therapeutic interventions. A more comprehensive understanding of the unique inflammatory response within the CNS following TBI could potentially reveal valuable therapeutic targets. By mapping specific molecular pathways, researchers can develop precise interventions and more effective treatments, tailored to the distinct challenges posed by different conditions
Comparative oxidation behavior of columnar and equiaxed Cr2AlC coatings
No full textThe oxidation behavior of columnar and equiaxed Cr2AlC coatings in air at 990 °C is studied. During the transient heating-up period to 990 °C the two coating morphologies exhibit different oxidation behavior. In columnar coatings, faster Al diffusion along column boundaries, triple junctions, and intercolumnar pores leads to the formation of metastable alumina polymorphs, resulting in rapid oxidation. In contrast, equiaxed coatings exhibit slower oxidation kinetics due to limited Al mobility, leading to direct formation of passivating α-(Al,Cr)2O3 through oxidation of Cr and Al at the surface
Tutorial Styles in Multiplayer Games : A Case Study of the Game Haunted Heist
No full textIn this study, two types of tutorials, referred to as a separate tutorial and an embedded tutorial, were studied and compared in order to find their strengths and weaknesses in onboarding players to networked low-complexity competitive multiplayer games. By utilizing our own game, Haunted Heist, we used practice-based research and developed these two types of tutorials ourselves in order to perform a qualitative A/B test, which gave us insights into when and why one tutorial type may be preferred over the other. This is beneficial to know, as tutorials in multiplayer games can be quite disruptive and prevent new players from joining multiplayer sessions with their peers, so exploring different types of onboarding to find methods that are more tolerated by players can potentially be very useful. The main findings of this research were that while separate tutorials are more consistent in their ability to teach, and generally give players a deeper understanding of the game’s mechanics, embedded tutorials can still reach a comparable level of understanding while simultaneously taking less time for the players to complete - meaning they are able to join their friends in play faster. Our findings suggest that embedded tutorials, which generally seem to be underutilized in multiplayer games, can be a viable onboarding method for these types of games
Euclid : II. The VIS instrument
No full textThis paper presents the specification, design, and development of the Visible Camera (VIS) on the European Space Agency's Euclid mission. VIS is a large optical-band imager with a field of view of 0.54 deg(2) sampled at 0 ''.1 with an array of 609 Megapixels and a spatial resolution of 0 ''.18. It will be used to survey approximately 14 000 deg(2) of extragalactic sky to measure the distortion of galaxies in the redshift range z = 0.1-1.5 resulting from weak gravitational lensing, one of the two principal cosmology probes leveraged by Euclid. With photometric redshifts, the distribution of dark matter can be mapped in three dimensions, and the extent to which this has changed with look-back time can be used to constrain the nature of dark energy and theories of gravity. The entire VIS focal plane will be transmitted to provide the largest images of the Universe from space to date, specified to reach m(AB) >= 24.5 with a signal-to-noise ratio S/N >= 10 in a single broad I-E similar or equal to (r + i + z) band over a six-year survey. The particularly challenging aspects of the instrument are the control and calibration of observational biases, which lead to stringent performance requirements and calibration regimes. With its combination of spatial resolution, calibration knowledge, depth, and area covering most of the extra-Galactic sky, VIS will also provide a legacy data set for many other fields. This paper discusses the rationale behind the conception of VIS and describes the instrument design and development, before reporting the prelaunch performance derived from ground calibrations and brief results from the inorbit commissioning. VIS should reach fainter than m(AB) = 25 with S/N >= 10 for galaxies with a full width at half maximum of 0 ''. 3 in a 1 ''.3 diameter aperture over the Wide Survey, and m(AB) >= 26.4 for a Deep Survey that will cover more than 50 deg(2). The paper also describes how the instrument works with the Euclid telescope and survey, and with the science data processing, to extract the cosmological information.For complete list of authors see http://dx.doi.org/10.1051/0004-6361/202450996</p
Best-practice guidance for Earth BioGenome Project sample collection and processing : progress and challenges in biodiverse reference genome creation
No full textThe Earth BioGenome Project has the extremely ambitious goal of generating, at scale, high-quality reference genomes across the entire Tree of Life. Currently in its first phase, the project is targeting family-level representatives and is progressing rapidly. Here we outline recommended standards and considerations in sample acquisition and processing for those involved in biodiverse reference genome creation. These standards and recommendations will evolve with advances in related processes. Additionally, we discuss the challenges raised by the ambitions for later phases of the project, highlighting topics related to sample collection and processing that require further development