<p>In this project a real-time face detection and recognition system has been discussed and implemented. The main focus has been on the detection process which is the first and most important step before starting with the actual recognition. Computably intensive can give good results, but at the cost of the execution speed. The implemented algorithm which was done is project is build upon the work of Garcia, C. and Tziritas, but the algorithm accuracy is traded for faster speed. The program needs between 1-5 seconds on a standard workstation to analyze an image. On an image database with a lot of variety in the images, the system found 70-75% of the faces.</p>

<p>In this research project I try to answer how context sensitive mobile devices can be used to complement traditional foreign language learning in a classroom. My approach to answering this question is based on the goal of creating a prototype application called the “Personalized And context sensitive foreign Language training system for real Life situAtions using mobile devices” (PALLAS). To be able to develop PALLAS satisfactory I start by performing a prestudy to create a theoretical ground for my ideas. This allows me to come up with ideas of how PALLAS could be used. To systemize my process I create a scenario-based development method. I start by writing user scenarios of the PALLAS system in use. This is the part where new ideas are included into PALLAS. Then I analyze the user scenarios and create formalized models of how PALLAS is supposed to work. The models are in turn used to design a system architecture, which finally is used to implement the PALLAS prototype. I finish my research by comparing the finished PALLAS prototype to the original user scenarios. It turns out that the process has resulted in an architecture that can almost completely satisfy the features found in the user scenarios. With this solid architecture in place the next natural step for PALLAS will be testing in real-life scenarios with students and teachers using the system.</p>

<p>With more and more cellular phone handsets being introduced wifi interfaces, an alternative to using the purely cell-based mobile phone networks infrastructure can be envisioned: mobile phone cells can be extended via ad-hoc relaying of traffic from out-of-coverage handsets towards in-coverage handsets, and mobile phones might be able to establish communication paths between them without passing through the cellular infrastructure. This presents a number of challenges including billing, security, multi-fabric routing, hand-over and quality of service. One task of this project is thus to look at the necessity of providing brokering network access, addressing variables influencing its success. The other task for this project is to construct the necessary mechanisms for establishing the ad-hoc part of an extension to cellular networks. This involves analysing ad-hoc networking and implementing OLSRv2, one of several experimental solutions for ad-hoc networking that exist within the Internet Engineering Task Force (IETF). To emulate such a system, but also for debugging and evaluating the implementation, a test-bed is required to be constructed.</p>

<p>The KOPEK Payment System supports payment of online content or services as well as single sign in. Setting up the KOPEK Payment System is easy, the system consist of the KOPEK Payment Server as well as an end user client. The server is installed so that it will be exposed to the desired customers, usually directly on the Internet. It can be installed in front an existing web server that handles more advanced content or services than the KOPEK Payment Server itself supports. Access to the content or the service will then be controlled by the rules configured in the KOPEK Payment Server. The KOPEK client application will be used from any end user who will access that content in order to identify the end user and make sure the user has met the requirements defined in the KOPEK Payment Server. Simple license control over personal licenses can be achieved by creating a web page and make the KOPEK Payment Server control access to that web page. An application can then try to access that web page and will only proceed if the web page can be read. If the web page cannot be read, the end user hasn’t met the requirements defined in the KOPEK Payment Server and hence the end user doesn’t have a valid license. This results in support for quite advanced copy protection when using the KOPEK Payment System. The copy protection can be further enhanced by also using the KOPEK Payment Systems support for centralized electronic identification systems, like the Norwegian BankID or BuyPass. In the market there exists some solutions that have specialized in license control and copy protection. These solutions seem to cover most of the more advanced licensing models, but none of these solutions can supply both a complete solution for license control and copy protection as well as a solution for electronic distribution of the licenses. There would probably be a market for such complete solutions and since the KOPEK Payment System already support both electronic distribution of licenses and can be used for simple license control an expansion of the KOPEK Payment System to also support more advanced license control. However, the KOPEK Payment System has some obvious weaknesses when it comes to dependence on an Internet connection and that both the company responsible for the KOPEK Payment System central servers as well as the company responsible for the KOPEK Payment Server used for license control can continue to keep the servers operative.</p>

<p>This thesis focuses on a part of Internet forensics concerned with determining the geographic location of Internet hosts, also known as geolocation. Several techniques to geolocation exist. A classification of these techniques, and a comparative analysis of their properties is conducted. Based on this analysis several novel improvements to current techniques are suggested. As part of an earlier designed Multi-Agent Framework for Internet Forensics (MAFIF), an application implementing two active- measurement geolocation techniques is designed, implemented and tested. Experiments with the application are performed in the Uninett network, with the goal of identifying the impact of different network properties on geolocation. What most clearly set this thesis apart from earlier work, in addition to the use of a multi-agent system, is the analysis of the impact of IPv6 on geolocation, and the introduction of multi-party computation to geolocation. The extensive focus on delay measurements, although not bringing anything new to the field of networking in general, is also new to geolocation as far as we know.</p>

<p>Power consumption becomes more important as more devices becomes embedded or battery dependant. Multipliers are generally complex circuits, consuming a lot of energy. This thesis uses Sand's multiplier generator, made for his master thesis, as a basis. It uses tree structures to perform the multiplication, but does not take power consumption into account when generating a multiplier. By adding power optimization to the generator, multipliers with low energy consumption could be made automatically. This thesis adds different reduction tree algorithms (Wallace, Dadda and Reduced Area) to the program, and an optimal algorithm might be found. After the multiplier tree generation, an optimization step is performed, trying to exploit the delay and activity characteristics of the generated multiplier. A simplified version of Oskuii's algorithm is used. To be able to compare the different algorithms with each other, a pre-layout power estimation routine was implemented. The estimator is also used by the post-generation optimization. Since accuracy is important in an estimator, the delay through a multiplier was also investigated. Taking the previous mentioned steps into account, we are able to get a 10% decrease in overall power reduction in a 0,18/0,15um CMOS technology, reported by "IC Compiler". Delay characteristics of a multiplier is also supplied, and can be used by other power estimators. This thesis shows how to achieve less power consumption in multipliers. It also shows that the delay model is important for estimation purposes, and how an estimator is used to optimize a multiplier. The findings in this thesis can be used as is, or be used as a basis for further study.</p>

<p>This study continues the work from the specialization project, carrying out the empirical study of the educational platform It's Learning at the Norwegian University of Science and Technology (NTNU). The goal of this study was to measure the user acceptance of It's Learning at NTNU, performing more surveys with the students at NTNU, but also with other educational establishments for comparison. In addition to this, we wanted more focus on It's Learning as a multi-user system. To achieve the goal we did a study of technology acceptance in information systems. In addition to this, we have studied important challenges of developing, introducing and using groupware to get more focus on It's Learning as a multi-user system. In addition to the survey we performed during the specialization project, three new surveys have been performed in this study. Two surveys were performed with student groups at NTNU, while one survey was performed with a student group at Sør-Trøndelag University College (HiST). In addition to this, we have performed interviews with some of the teachers at NTNU to get an indication of their usage and opinion of It's Learning. Instead of using one of the existing technology acceptance models as a framework for the survey, we created a new research model. The new model combines elements from two existing technology acceptance models, and was created to better adapt to It's Learning. The survey, together with the interviews with the teachers, and the groupware challenge study, created a basis for an evaluation of It's Learning. The results from the survey in this study clearly contradicted the results from the survey in the specialization project, but even though the results differed, all the student groups identified many of the same topics when it comes to problems with It's Learning and how to improve the system The results from the study indicates that two of the main problems with It's Learning at NTNU is that too few subjects use It's Learning, and that It's Learning is not utilized in a good way by the teachers. These two problems represent two important challenges with multi-user systems, the "critical mass" challenge and the "non-standard use" challenge. As opposed to NTNU, HiST has mandatory use of It's Learning, meaning that they avoid the "critical mass" challenge. We also found that It's Learning has an inflexible and cumbersome user interface, and the lack of possibilities for integrating the calendar and message system in It's Learning with other existing applications affect the utility value of these features. That It's Learning is a closed system is one of the largest problems for the teachers, in addition to some missing features for administrating exercises in their subjects. From this study we found that general user acceptance models give a good indication on the user's perception of an information system, but that there is also important to focus on the challenges related to specific types of systems, like the multi-user environment in the case of our study.</p>

<p>A research group at the Norwegian University of Science and Technology plan to integrate Capacitive Micromachined Ultrasonic Transducers (CMUTs) in a probe which can be used to image human blood vessels from the inside. With CMUT technology still being in its youth, it is important to characterize different CMUT prototypes to learn which design is best suited for the probe. In this thesis, a network analyzer and a heterodyne interferometer are used to characterize two prototypes, both electrically and optically. The heterodyne interferometer was proven to be extremely sensitive and thus a valuable tool when investigating the behaviour of single CMUT cells. Measurements revealed that the cells within both the CMUT arrays had different resonance frequencies. This effectively broadens the band widths of the prototypes, but reduces their maximum acoustical output power, given that all the cells operate at the same frequency. It was found probable that the CMUT characteristics were influenced by how charges moved within the prototypes. These charging effects are quite complex and would in practice make it hard to predict where the centre frequency of the CMUT array lies from time to time. The vibration pattern of most the cells at their respective resonance frequencies looked healthy both in air and immersed in a liquid when measuring with the interferometer. Immersed, the resonant frequency range of the CMUT become wider, but the peak was located at a lower frequency than in air. The next CMUT prototype to be tested in the future does not fit into the experimental setup. A plan to adapt the setup has been presented in this thesis, making it more practical to handle different samples. This design would also make it easier to investigate crosstalk effects between neighbouring CMUT elements within the arrays. Knowledge aquired in this thesis will hopefully help researchers figure out which areas of the CMUT design have the potential to be improved.</p>

<p>In this thesis a feasibility study of a direct AC-AC converter for wind power application has been performed. Two three-phase voltages are converted to a single phase square wave for input to a high-frequency transformer and then rectified. No DC-capacitor is present in the converter and bi-directional switches consisting of two reverse blocking IGBTs allow the direct AC-AC conversion. Efficiency, silicon usage, price and output quality of the converter were to be compared with a reference case and conclusions to be drawn from the obtained results. The price comparison could not be finalized due to lack of price data. The efficiencies were found to be 97:7 % for both the converter setups while the silicon usage was 4:6 % lower for the direct AC-AC solution. A back-to-back converter was used as the reference case. The harmonic content was less in the back-to-back converter and DC-offsets in the phase currents caused power oscillations for both converters but they were higher in the direct AC-AC converter. The reference case needs more switches than the direct AC-AC solution if put in a split drivetrain configuration and DC-capacitors are also present in the former. The size is therefore expected to be lower for the AC-AC solution. The obtained results was used to conclude that the direct AC-AC solution is a feasible solution for a split drivetrain configurated wind turbine with multiple generators. The voltage of the square wave in the direct AC-AC converter needs to be two times higher than the DC-voltage in the reference case to obtain the same rated voltage in the generators when they are supplying power simultaneously. The switch voltage ratings must then be increased accordingly. Another application was proposed to better exploit the converter topology examined and to avoid the doubled voltage rating. An isolated power system with a time varying energy source such as a wind turbine needs an auxiliary power source to be able to supply the full load power when the other is unavailable. The double input converter can then control the two generators so that a constant output power is achieved. The square wave voltage amplitude can then be dimensioned from the load power only rather than for the total rated power in the two generators. It is therefore halved and the voltage stresses on the switches are similarly reduced.</p>

<p>En av de dominerende IT-avgjørelsene tatt av organisasjoner gjennom det siste tiåret, har vært valgene i retning av delte informasjonssystemer og databaser. Integrasjon av separate og isolerte egenutviklede systemer, som begynte som prosjekter rundt noen få applikasjoner, har i dag fortsatt i en mye større skala gjennom initiativ som standardiserte virksomhetssystemer og datavarehus. Selv om nåtidens perspektiv er preget av antagelsene om avansert teknologi som enten determinanter eller muliggjører for radikale organisatoriske endringer, viser empiriske studier en rekke uoverensstemmelser i tilfeller som støtter en slik deterministisk tankegang. Samtidig erfarer verden i disse dager en ny periode som har fått betegnelsen kunnskapsalderen. I denne nye sammenhengen er kunnskap en primær råvare og kunnskapsflyt er ansett til å være den viktigste faktoren i økonomien. Siden hurtige teknologiske innovasjoner bygger broer mellom konkurrerende organisasjoner er det nå en trend i industrien å betrakte den felles kunnskapen til de ansatte som en nøkkelfaktor til å produsere innovative og konkurransedyktige produkter. Denne endringen i fokus medfører at organisasjoner må revurdere måten de utfører sin forretning på siden fokus ikke lenger er på materielle aktiva, men på mennesker sine evner og erfaringer. Denne rapporten ønsker å bedre forståelsen av virksomheter sine egentlige behov for kunnskap og kunnskapsoverføring, med utgangspunkt i at det som i mange tilfeller er antatt å være formell praksis ikke er den samme som faktisk er utført. Dette inviterer til diskusjon om hvorvidt standardiserte virksomhetssystemer basert på beste praksis i industrien vil kunne imøtekomme de lokale kravene som stilles i organisasjoner. I lys av teori om praksisfellesskap og vurderte kritiske suksessfaktorer basert på tidligere erfaringer med implementering av virksomhetssystemer søker vi en bedre forståelse for eksisterende praksis i organisasjoner for å bedre kunnskapsoverføringen. Det er i den anledning blitt foretatt en case-studie hos salgsrepresentantene i Felleskjøpet Trondheim. Nøkkelord: ERP, CRM, kunnskapsforvaltning, praksisfellesskap, teknologi i praksis</p>