In this project, the growth parameters of Pulsed Laser Deposition are optimized in order to obtain a high quality interface between two complex oxides, LaAlO3 and SrTiO3. The prepared samples are compared by their sheet resistance, and the influence by the various growth parameters are investigated.The main finding is that reducing the laser fluence significantly lowers the sheetresistance. The lowest obtained sheet resistance was 81.2kohm. This value wasobtained with a laser fluence of 0.7 J/cm^2 and the results indicate that morecan be gained by going even lower. In previous work, a strong anisotropy of the sheet resistance has been observed. The anisotropy is further studied in this project and compared with the step-and-terrace topography at the interface. Atomic Force Microscopy is used to obtain the step directions and terrace widths, and these values are compared with resistance measurements. No correlation is found between these data. The results suggests that the anisotropy is due to other factors than the steps-and-terraces.

In this thesis, methodology for partial self-reconfiguration of synchronous modules has been developed. A simple software-based scheduler has been built for scheduling synchronous modules on the FPGA. The motivation behind this was that partial reconfiguration of synchronous modules at run-time had not been performed earlier in the AHEAD-project. Also, the project report written by the same author as this thesis has shown that a synchronous module can be replaced in a bitfile. However, the project report did not perform this reconfiguration at run-time.Based on the project report, the problem has been decomposed and simple tests using clocked flip-flop designs have been performed on the FPGA. These tests forms a proof-of-concept for partial self-reconfiguration of synchronous modules on the Virtex-4 FPGA. However, the tests also showed that the reconfiguration time was quite high. It took several seconds to write one partial bitstream to the configuration memory.Vegard Endresen has previously made a backend module for data transfer between the HWOS and a reconfigurable module. Experiments were performed in this thesis to see if the clocking methodology could be integrated into this backend module. The module could be built with the methodology, but a running solution on the FPGA was not shown.The software part of the HWOS was rewritten from scratch as the previous version was not thoroughly analyzed. A round-robin scheduler using priority queues has been implemented. A test-driven development technique has been used for development, hopefully making the system more robust. The scheduler is a part of a daemon running on the embedded system, where a message server handles requests for new processes and a placer places new tasks on the FPGA. The complete system was initially based on ideas and code developed by Sverre Hamre and Vegard Endresen in previous AHEAD-projects.

Testing and verification of digital systems is an essential part of product develop-ment. The Network on Chip(NoC), as a new paradigm within interconnections;has a specific need for testing. This is to determine how performance and prop-erties of the NoC are compared to the requirements of different systems such asprocessors or media applications.A NoC has been developed within the AHEAD project to form a basis for areconfigurable platform used in the AHEAD system. This report gives an outlineof the project to develop testing and benchmarking systems for a NoC. The specificwork has been regarding the development of a generic module connected to theNoC and capability of testing the NoCs’ properties. The test system was initiatedby Ivar Ersland in 2009 and developed further by Andreas Hepsø, and MagnusNamork in the fall of 2010. The functionality and systems that are implementedare the following:• Fully functional Hardware/Software interface which defines communicationbetween NoC the user• Reactive system which responds to interaction based on package information• MPEG example system that mimics an MPEG data stream• Software reconfiguration of the traffic tables by sending specific packages tothe system• Cell processor example application to test simple computation and commu-nicating modules on the networkThe systems have been tested successfully, verified and implemented on a XilinxSpartan FPGA. It has also been developed a software system written in C to read and interpret data from the Network in on-chip tests. In total these imple-mentations have been the foundation of building a benchmarking platform for theNoC.

Det finnes flere gode måter for personer å forebygge og motvirke hørselskader på. En av disse metodene er ved å måle lyddosen en person blir utsatt for. Denne metoden er grunnlaget for dette prosjektet. Prosjektet ”Distribuert lydnivåmåling” har resultert i en prototype av et system som skal varsle personer som oppholder seg i miljøer med farlige lydtrykk. Prototypen består av to sensorer som sender data til en datamaskin som er mottakeren i systemet. Hver sensor består av mikrofon, filter, mikrokontroller og radiomodul. Sensorene måler momentant lydtrykk og peak-verdi som mottakeren visualiserer. Det ekvivalente lydnivået blir også utregnet av mottakeren og vist sammen med de andre parametrene på en oversiktlig måte. Kravene til en dosimetrimåler er definert i en egen standard, og det er en lydtrykknivåmåler som i tillegg beregner det ekvivalente lydtrykket måleren blir utsatt for over tid. Denne standarden definerer blant annet frekvensrespons, batterilevetid og dynamisk område. Underveis i prosjektet har det blitt jobbet mye med å finne gode løsninger på kravene til prototypen fra denne standarden. Det ble designet to sensorer i prosjektet da det viste seg at den ene ikke tilfredsstilte standardens krav til frekvensrespons. Det viste seg også at begge disse sensorene hadde analoge kretser som ga for mye støy. Dette gjorde at støygulvet i kretsene ble for høyt og senket det dynamiske området til enhetene. Batterilevetiden og bruk av strømgjerrige komponenter i prosjektet ble sett på som viktig, og prototypen har en teoretisk lenger batterilevetid enn standarden krever. Ved prosjektets slutt er det ferdige produktet to sensorer som måler det meste av det ønskede dynamiske området med god nøyaktighet. Dataene blir sendt til en mottaker, hvor det ekvivalente lydnivået blir beregnet og vist grafisk. Produktet er også skalerbart, og kan bestå av opptil 90 sensorer som kan rapportere til samme mottaker.

In this thesis, methods for automatic detection for radar systems are investigated. The objective is to indicate the presence of targets in the midst of noise and clutter. One of the most efficent methods for doing this is to exploit the Doppler shift in reflections from moving targets. This is called Moving Target Indication (MTI), and it is used in many radar applications today. However, such functionality is not typical for radars employing a magnetron oscillator. The magnetron oscillator is widely used in civil marine radars, and MTI processing is of interest for such radars as well. In addition to MTI processing, automatic detection may be applied in order to make decisions on target presence. This may be achieved by employing Constant False Alarm Rate (CFAR) detection and pulse integration. The challenges with automatic detection are prediction of the clutter power, and handling of non-homogeneous environments. Ideally, clutter components will be removed by the MTI process, leaving receiver noise and reflection from targets at the output. However, this is not necessarily the case when applied to a magnetron radar. A particular automatic detector employing Ordered Statistics (OS) CFAR and binary pulse integration is investigated. This is a robust detector that may operate in the presence of multiple targets and non-Gaussian clutter. The binary integrator contributes in reducing the false alarm rate, such that acceptable performance may be achieved. In order to find suitable parameters for the detector, analysis of an MTI processed radar signal containing reflection from waves at sea and a small boat is carried out. Also, the performance of the detector has been measured in terms of false alarm probability, and target detection. Analysis shows that Weibull or K-distribution are suitable models for the sea clutter, and that the MTI signal exhibits spatial correlation between clutter samples. The correlation is concluded to be the reason for degradation in performance, as detection on clutter appears as targets. Also, optimum parameters for the detector is found, and it is shown that increasing the number of reference samples increases the number of target detections.

In the work with this master thesis, electron beam lithography (EBL) and other cleanroom methods was used to develop a reproducible process for making ohmic contactsto nanowires (NWs), with the aim of investigating different dopants’ influence ontheir physical properties. After optimizing the EBL process, good alignment of thecontact patterns to the NWs was achieved by doing a thorough alignment procedure.A detailed write field alignment close to the sample position prior to the 3-point alignment, along with consistent use of alignment positions, was found the be especially important to reduce misalignment.For surface treatment it is shown how a diluted HCl solution is used to removenative oxide on the NW surface, and how a solution of citric acid and hydrogen peroxidecan be used to effectively etch the AlGaAs shell of core-shell nanowires, thoughselective etching of AlGaAs on GaAs is difficult. Further, an ammonium sulphidesolution is used to passivate the surface prior to metal deposition.Metalization was done using an electron beam evaporator, however optimal thinfilms could not be achieved due to the presence of strain. This is believed to originate in the evaporation of electron beam resist (ER) during metal deposition due to high local heating of the samples. While this was remedied by, among other, depositing thinner films at high deposition rates, it could not be fully avoided. Nevertheless, satisfactory lift-off was still achieved in most cases.IV characterization of the different samples was done using a probing station. Thesamples generally display asymmetric and non-linear characteristics, indicating thatohmic contacts have not been achieved. Annealing tests was also done on a numberof samples, and while this generally resulted in more linear IV curves, the currentwas also reduced in most cases.

This thesis presents a realization process, of how a programmable DDS/PLL signalsource were made suitable as an linear FMCW waveform generator for automotiveW-band radar applications. The dissertation describes the specific background theorythat were used to complete the implementation and substantiate the discoveries duringthe development process. Hence is an outline of FMCW basic properties, automotiveradar implementations, applied waveforms and the fundamental radar ambiguity functionpresented and threated. Moreover are vital background theory for basic FMCW designproblems presented, including degradation of range resolution due to loss of effectivebandwidth, nonlinearities in chirp modulation and limitations of the receiver resolution.Additionally is the inevitable problem of FMCW transceiver noise leakage briefly described, along with the general concept of choosing the right beat frequency for maximal FMCW sensitivity and maximal range-Doppler resolution. The specific circuit design is then presented, together with the full radar design which the signal source is intended for. The realization process starts with an initial test of circuit performance, seen in relation to the radar documentation and set the fundament for the further investigation and development. Hence does this part include measurements and discussion of output power achievements, spectral purity, spectral waveform appearance and phase noise. The thesis does then set further focus to more specific methods of measuring and evaluating the circuits LFM waveform, in aspect of a combined frequency and time characterization, chirp linearity achievements and waveform beat frequency evaluation. The dissertation does further describe how the first hand waveform generation were erroneous and how this problem was investigated and solved through radar documentation/source code corrections. As a last part of the realization process is then the final and successive implementation of waveforms described, along with the main results presented as waveform spectrograms and selected beat frequency spectrum plots. The working process and the main results are then summarized in a final summary discussion. The major results and conclusions of the thesis are that the tested NORBIT DDS/PLLsignal source could be realized as an linear FMCW waveform generator with overallgood linear properties. Two basic LFM waveforms, the up-chirp and triangular waveformwere implemented and tested. However were there discovered nonlinearities withinthe up-chirp waveform, due to the transient response of the total circuit. Hence mostlikely caused by the limited PLL frequency lock time. This was proven through aspecific beat frequency analysis of the most affected waveform, with fitted W-band LFMcharacteristics. Nevertheless were the isolated chirp generation within the waveformsconsidered more than sufficient, since both windowing technics and selective samplingcould be used in the future radar implementation. Three specific waveform realizationswere specially recommended for the initial further use. Further were also a specific region of desirable waveform chirp-rates recommended, that enables a good trade off between close target radar sensitivity and digital acquisition system complexity. Additionally did the thesis work conclude with that the NORBIT signal source design, enabled a flexible and easy control of the waveform generation through the microcontroller interface. Further more had also the thesis work resulted in the implementation of two good methods for waveform measurements and analysis. By the use of the spectrogram- and the beat frequency-method, could accurate waveform properties be extracted. Hence were also these methods recommended for further waveform developments in the experimental radar project.Furthermore was it recommended that future effort should be put in to the developmentof more applicable waveforms, to achieve full multiple range-Doppler target extraction.Hence since the overall signal source proved to have the nessesary waveform agility.It was in addition proposed that the future full radar implementation should utilizethe accessabillity of the microcontroller to achieve synchronization of signal sourcemodulation and future sampling solution. Thus to enable beneficial beat frequencysampling for the quadrature radar receiver. Since the mapping of the phase noiseinfluence and the chirp linearity were seen as non-optimal, was it additionally suggested that future studies should yield better methods for such characterization. It was finally put forward that an intermediate simple test radar could be implemented to enable sampling system testing and FMCW signal processing of real measurements, if the millimeter partof the radar is further postponed.

Mange av dagens trådløse systemer benytter frekvenser opp i titals gigahertz området,ofte kalt RF-bånd, til å sende bredbåndede RF-signaler. Siden den digitale enheten somprosesserer RF-signalene opererer på et lavere frekvensområde, må RF-signalet konverteres til og fra et såkalt basebånd. Superheterodyn-prinsippet er per dags dato den vanligste frekvenskonverterings-metoden til og fra basebåndet, men metoden begrenser ofte systemytelsen og tar stor plass. Av den grunn har denne masteroppgaven sett påSampling Down Conversion (SDC) og Sampling Up Conversion (SUC) som et alternativ for å flytte bredbåndede RF-signaler henholdsvis ned og opp i frekvens.Prinsippet og ytelsen til SDC og SUC er blitt analysert gjennom en studie av bakgrunnsteorien, simuleringer i MATLAB og praktiske laboratorieforsøk. Bakgrunnsteorien har tatt for seg en ideell beskrivelse av SDC og SUC, hvor prinsippet og komponentene som inngår i frekvenskonverteringen har blitt gjennomgått. Prinsippene blevidereført til MATLAB, hvor SDC og SUC har blitt simulert ved å flytte et RF-signal tilog fra et basebånd. Simuleringene tok utgangspunkt i det ideelle tilfellet, hvor enkeltebegrensninger påpekt i bakgrunnsteorien har blitt innført. For å verifisere simuleringeneble det utført praktiske laboratorieforsøk, hvor SDC og SUC har blitt demonstrert på eteksempelsystem med begrenset ytelse.Gjennom analysen er det blitt vist hvordan SDC konverterer et bredbåndet RF-signalned til basebåndet, ved å punktprøve direkte med en analog-til-digital-omformer (ADC).Resultatene viser hvordan kravet til ADCens analoge båndbredde øker, og hvordan SDCer ekstra følsomt for fasestøy i form av jitter. I oppkonverteringen fra basebåndet haranalysen vist hvordan SUC utfører frekvenskonverteringen, ved å båndpassfiltrere direk-te etter en digital-til-analog-omformer (DAC), og hvordan valget av DAC-type påvirkerytelsen. Resultatene viser hvorfor en return-to-zero (RTZ) DAC bør benyttes, og hvordan punktprøvingsfrekvensen påvirker frekvenskonverteringen. For både SDC og SUCviser resultatene hvordan båndpassfiltre med gode frekvensegenskaper, er helt sentraltfor ytelsen i ned- og oppkonverteringen.Siden frekvenskonverteringen foretas direkte på en ADC og DAC, føres den digitale enheten nærmere antennen, og den faktiske størrelsen minker i forhold til superheterodyn-prinsippet. En annen stor fordel er at fleksibiliteten øker ved at frekvenskonverteringen styres direkte av punktprøvingsfrekvensen. Hovedkonklusjonen til oppgaven er at SDC og SUC kan benyttes til frekvenskonvertering, samtidig som fleksibiliteten øker. Hvilket frekvensområde SDC og SUC kan operere under, er helt avhengig av kvaliteten på komponentene som benyttes. For å opprettholde akseptable ytelser har oppgaven vist hvordan det stilles strenge krav til komponentene som inngår i frekvenskonverteringen.

The increased processing power and screen sizes of mobile devices has made it desirable to watch multimedia presentations on the go. On such devices the data network bandwidth is usually the limiting factor, which imposes a tradeoff between quality and resolution on the presented content. A new video compression system called High Efficiency Video Coding (HEVC) is currently under development. The vision of HEVC is to create a compression system that achieves the same quality at half the bit rate compared to the existing H.264/AVC standard [2].The goal of this thesis is to investigate how HEVC performs compared to H.264/AVC using mobile platforms and sport content as the scenario. The subjective test was conducted on an Apple iPad. It indicated that HEVC has a clear gain in compression compared to H.264/AVC. On average at a resolution of 640x368, HEVC achieved a good quality rating at approximately 550 kilobit per second while H.264/AVC did almost reach this quality at 1000 kilobit per second. However, it was shown that subjective quality gain varied over content.The objective measurements showed an overall reduction in bit rate of 32% forthe luma component. However, the reduction of bit rate was highly variable over content and resolution. A high correlation between the subjective and objective measurements was found, which indicates that it was almost a linear relationship between the reported subjective and objective results.In addition, a proposed deblocking filter was implemented. The filter applies a new filter function of the luma samples and performs line based filtering decision. On average the reduction in bit rate was reported to be 0.4%, with a maximum reduction of 0.8% for the luma component. The decoding time relative to the second version of the HEVC test model was reported to be 1.5% higher. This is most likely due to the line based filtering decision. The general impression of HEVC is that it has the ability to reach the stated vision, and perhaps even surpass, when finalized.

In broadcasting environments intra-only video coding with JPEG 2000 has shownto provide many desired features along with high picture quality. However, thereis no exploitation of temporal redundancy, which can reduce the bit rate whilemaintaining the quality.Optical flow algorithms are designed to find the apparent movement of brightness inan image, and can be used to estimate the motion of each pixel between consecutiveimages. This thesis explores the performance of a hybrid video codec that uses the’Classic+NL’ optical flow algorithm for motion compensation and JPEG 2000for encoding the estimation error.The motion estimation proved to be inaccurate at the edges of objects. This cancause high frequency components in the residue image, which will decrease theefficiency of JPEG 2000. Occluded regions will also have poor estimation, as theyare not present in the previous frame. Since noise is not considered when the opticalflow is calculated, the energy of the noise may increase after motion compensation.Without addressing these issues, optical flow algorithms are not well suited formotion estimation in hybrid video codecs.Even with the inaccurate motion compensation performed by the optical flow algorithm,there was an overall reduction in bit rate of 18.8%, compared to intra-onlycoding with JPEG 2000. The performance was highly content dependent, rangingfrom a reduction of 90% to an increase of 27%. The reduction comes at a cost ofincreased delay, higher complexity, vulerability to transmission errors, and a lackof a constant bit stream. The proposed hybrid codec is therefore not suited toreplace intra-only coding with JPEG 2000 in contribution environments.