Energy consumption and correlated greenhouse gas emissions are a big global problem. It affects all parts of society, and each industrial sector must work toward reducing itscarbon footprint. This thesis details the research of different methods to model the energyconsumption of video streaming, and works towards creating a final model. The videostreaming process is broken down into a core process consisting of head-end, distribution,transmission, and terminals. The process that contributes the most to energy consumptionat the head-end is found to be video encoding. This thesis explores the energy consumption of video encoding in depth and how it is affected by parameters such as hardware,codec choice, codec preset selection, and video details such as resolution, framerate, andduration, but these parameters are found to be insufficient to accurately model the energyconsumption of video encoding. In distribution and transmission, the highest contributor is found to be content delivery networks. The energy consumption of content deliverynetworks is investigated however no appropriate model is found. For terminals, the mostimportant factor is the kind of terminal used. The energy consumption of televisions, desktop computers, laptops, and mobile terminals is investigated, and models are presented foreach. The thesis also discusses the different models, their advantages, and their shortcomings. Additionally, an application to visualize features of the model is presented.

There has been extensive research and progress on virtualization on CPUs for a while. More recently the focus on GPU virtualization has increased as processing power doubles roughly every 2.5 years. Coupled with advances in memory management and the PCIe standard the first hardware assisted virtual solutions became available in the 2010s. Very recently, a new virtualization mode called Multi-Instance GPU (MIG) makes it possible to isolate partitions with memory in hardware rather than just software. This thesis is focused on virtual GPU performance and capabilities for AI training in a multi tenant situation. It explores the technologies currently used for GPU virtualization,including Single Root IO Virtualization (SR-IOV) and mediated devices. It also covers a proposed new standard for IO virtualization called SIOV that addresses some of the limitations in the SR-IOV standard. The limitations of time sliced virtualization are mainly the lack of customization for a partition compared to CPU virtualization and the problem of overhead. MIG virtualization is more customisable in how compute power and memory can be allocated, the biggest limitation is that fast intercommunication is not currently possible between partitions, making MIG more suited for applications that can run on just one partition. It is also not suited for graphical applications as it currently does not support any graphical APIs. The experimental results showed that in compute situations the overhead of time sliced virtualization is around 5% while the maximum intercommunication bandwidth is lowered by 11% and latency increased by 25%. Time slice windows of 4ms compared to 2ms can decrease scheduling overhead to nearly 0.5% at the cost of increased latency for the end user, this can be beneficial for applications where user interactivity is not of importance.

Along with the increased popularity of browsing and purchasing clothes in online stores, thelogistical and environmental issue of returning unwanted items has reached alarming proportions.In an attempt to lower the amount of returns, a range of companies are now considering ways inwhich the online shopping environment can provide a more immersive and informative experienceakin to a physical store – a concept called virtual try-on. In this report, visual and interactiveshortcomings of existing solutions are addressed by approaching the areas of virtual outft representation and creation from a video game-oriented point of view. Additionally, a framework fortranslating the concept of character customization from the video game world to the virtual try-onworld is explored with the aim of providing means to enhance customer engagement. Throughdeveloping and evaluating a proof of concept implementation, we propose a method for creating asimple, yet fexible, outft creation system, as well as techniques and considerations for rendering3D apparel items in a performant, interactive, and realistic manner. Finally, approaches for adapting character customization-related features to virtual try-on are proposed through psychologicaltheories by considering the personalized virtual avatar as the central element. / I samband med ökad nätklädeshandel har antalet returnerade produkter nått alarmerande proportioner, vilket har både logistiska och miljömässiga implikationer. I syfte att minska antalet returneringar inom nätklädeshandeln utvecklas just nu olika metoder ämnade till att ge kunden verktyg till att utvärdera produkter online på ett sätt som liknar de traditionella provhyttarna – ett concept som kallas virtual try-on. Denna rapport tar sig an visuella och interaktiva brister som återfinns i nuvarande lösningar genom att angripa problem relaterade till framställningen och sammansättningen av virtuella kläder från ett datorspelsorienterat perspektiv. Dessutom utforskas ett ramverk för överförandet av konceptet character customization från datorspelsvärlden till virtual try-on i syfte att möjliggöra utvecklingen av mer engagerande virtuella miljöer. Genom att först utveckla och utvärdera en konceptuell virtual try-on implementation föreslår vi sedan en metod för att skapa ett enkelt, men samtidigt flexibelt, system för att skapa virtuella ytstyrslar, samt tekniker och överväganden kopplade till effektiv, interaktiv samt realistisk rendering av tredimensionella klädesplagg. Slutligen framförs tillvägagångssätt för att adaptera koncept relaterade till character customization till virtual try-on genom ett ramverk baserat på psykologiska teorier, där den personliga virtuella avataren står i centrum.

The era of modern computing has been the stage for numerous innovations that have led to cutting edge applications and systems. The characteristics of these systems and applications have been described and quantified by many, however such information is fragmented between various repositories of system and component information. In an effort to collate these disparate collections of information we propose iLORE, an extensible data framework for representing computer systems and their components. We describe the iLORE framework and the pipeline used to aggregate, clean, and insert system and component information into a database that uses iLORE's framework. Additionally, we demonstrate how the database can be used to analyze trends in computing by validating the collected data using previous works, and by showcasing new analyses that were created with said data. Analyses and visualizations created via iLORE are available at csgenome.org. / Master of Science / The era of modern computing has been the stage for numerous innovations that have led to cutting edge applications and computer systems. The characteristics of these systems and applications have been described and quantified by many, however such information is fragmented amongst different websites and databases. We propose iLORE, an extensible data framework for representing computer systems and their components. We describe the iLORE framework and the steps taken to create an iLORE database: aggregation, standardization, and insertion. Additionally, we demonstrate how the database can be used to analyze trends in computing by validating the collected data using previous works, and by showcasing new analyses that were created with said data. Analyses and visualizations created via iLORE are available at csgenome.org.

A user-centered approach to interactive system development requires a way to represent the behavior of a user interacting with an interface. While a number of behavioral representation techniques exist, not all provide the capabilities necessary to support the interaction development process. The original goal of this research was to modify and extend the User Action Notation (UAN), a user- and task-centered behavioral representation technique. In order to facilitate and evaluate the improvement in the UAN, we developed and evaluated a taxonomic model of behavioral representation techniques. The development and evaluation of our model followed the epistemological cycle of observation, theorization, and evaluation. The model provides a framework for discussing, analyzing, extending, and comparing existing behavioral representation techniques, as well as being a springboard for developing and evaluating new techniques. / Ph. D.

Denna rapport presenterar ett utvecklingsarbete av standardiserade tillvägagångssätt för att bygga ett ramverk för dokumentsignering. Studien fokuserar på tre olika delar av dokumentsigneringen. De tre delarna är identitetsautentisering, signering och verifiering av signering samt versionshantering. Efter att ha utfört en undersökning konstaterades det att den mest lämpliga lösningen för detta system är att använda X.509v3-certifikat som autentiseringsmetod för identiteter, tillsammans med en PKI( Public Key Infrastructure). För signering och verifiering så används digitala signaturer tillsammans med algoritmerna SHA-2 samt RSA, vilket är två välkända och populära tillvägagångsätt i dessa sammanhang. Andra tillvägagångssätt har diskuterats samt framtida utvecklingsmöjligheter och förbättringar. / This report presents an investigation into different standardized approaches to building a framework for document signing. The study focuses on three different parts of document signing. The three parts are identity authentication, signing and verification of signing and version management. Through the investigation of standards in the market, it was found that a solution using X.509v3 certificates as an authentication method for identities together with a PKI (Public Key Infrastructure) was best suited for this system. For signing and verification, digital signatures are used together with the algorithms SHA-2 and RSA, which are two well-known and popular approaches in these contexts. Other approaches have been discussed as well as future development possibilities and improvements. / Datateknik, Dokumenthantering, Säkerhet, Dokumentsäkerhet

Prevalent in dialogues are communications about the activity of talking itself. Examination of dyadic conversations readily shows the ubiquity of statements such as "Okay," "Well," "You know," indicating that in natural, spontaneous talk people are often concerned with the process of talking as well as with the content of their talk. In an analysis of a corpus of twenty seven transcribed telephone conversations (80 minutes of continuous talk). we found that such statements occur with an average frequency of 1.49 per conversational turn, and that a turn free of such statements is rare. A verbal expression focusing on the activity of talking we term a meta-statement. In this thesis, we present a model of meta-statements. We propose that metastatements are pragmatic units of talk used to regulate and maintain the conversational contexts shared by interactants. More specifically, we hypothesize that to determine the "meaning" of meta-statements, six aspects of a conversation's context must be modeled: the rhetorical, conversational, interpersonal. emotional, cognitive, and communicational aspects. We specify these models in what follows and outline the components of a conversational system that is able to parse the meaning of meta-statements. / Master of Science

Designing a User-System Interface (USn is a complex task and has been approached in many ways. One approach has been to use USI design guidelines to help improve the quality and consistency of USIs. To use guidelines effectively, a general set of guidelines must be tailored to a specific application. One popular set of guidelines is the Smith and Mosier Guidelines for Designing User Interface Software (1986). This thesis investigated the effects of using a computer-based hypertext design aid (DRUID, Dynamic Rules for User Interface Design) for the selection of USI guidelines for both experienced and novice Guidelines users. Also, the relative usability of the software and hardcopy document was examined to improve the interface for the next iteration of DRUID. Both performance and variability in guideline selection strategies were studied within the context of the experimental tasks. Results indicate that subjects effectively used both the book and hypertext presentation media. However, subjects selected more relevant guidelines with the book. The media influenced the subjects' selection strategies, and the subjects read less when using DRUID. Subjectively, the software was preferred because it provided assistance in the selection process. The presentation of the guidelines could be improved for both media, and many of the factors which led subjects to select fewer relevant guidelines with DRUID could be overcome with improved USI design and implementation. Also, to overcome the limitations of presenting large texts on-line, "value added" features should be incorporated into on-line tools to help users better perform their tasks. / Master of Science

This study used a standardized evaluation to compare two direct manipulation graphics interfaces. The Interfaces investigated were (1) object-oriented (vector) graphics and (2) bit-mapped graphics (object-oriented graphics interfaces are not to be confused with object-oriented programming or object-oriented data bases). Experienced and novice subjects performed objectively derived benchmark tasks appropriate for two-dimensional graphics packages. Both performance and preference data were collected. Task completion time, aborted attempts, learning effects, and errors constituted the performance measures. For the preference data, subjects completed questionnaires to rate the interfaces on both an absolute and a relative basis. Results indicate that the object-oriented graphics interface is superior to the bit-mapped interface. Both the performance and preference measures indicate that the object-oriented interface is better suited for manipulation of graphics. / Master of Science

The UAN (User Action Notation) is a task- and user-oriented behavioral representation technique for specifying interface designs that is used to communicate interface designs, support task analysis, and facilitate usability evaluation. A UAN interface description is made up of several linked sheets of paper containing task descriptions, notes, and sketches. UAN users have reported that writing and reading these designs is difficult. It is our hypothesis that these problems in usability are not inherent in the notation itself, but rather that effective use of UAN is hampered by the manual manipulation necessitated by pencil and paper. QUANTUM (Quick User Action Notation Tool for User interface Management), a software tool to facilitate writing and reading of UAN for user interface design, was developed through an iterative process based on formative evaluation by expert UAN users. QUANTUM supports development of a graph-based representation of a task abstraction hierarchy for a user interface design, with each node representing a user task. Task description windows for each task node allow entry of UAN in a tabular, spreadsheet-like format. Task libraries enable designers to build archives of task descriptions, enabling their reuse in other interface design projects. Note addenda containing explanatory text, screen sketches, design questions, and/or audio/video can be placed anywhere within a design description. These features, not afforded by pencil and paper, give QUANTUM a functional advantage over such media. Expert evaluators used in qualitative formative evaluation of the QUANTUM prototype indicated a strong preference over pencil and paper for using QUANTUM to produce UAN task descriptions. As a result of this work, we claim that QUANTUM improves usability of the UAN. / Master of Science