Thesis (MScEng)--Stellenbosch University, 2011. / ENGLISH ABSTRACT: A well-known problem with FM-CW radar systems is the leakage of transmitter power into the receiver which leads to the making of close-in targets, and can severely limit the system dynamic range performance. This thesis considers two solutions to this radar system problem for a low frequency radar operating in the VHF band. The first method to suppress coupling is using separate transmit and receive antennas designed in such a way as to reduce coupling between them. The second is to design a negative feedback loop as part of the radar receiver where the feedback loop adaptively reduces the amount of transmitter leakage through to the receiver. This project details the realisation of these two solutions. A number of antenna designs are modelled in software and simulated to determine their characteristics of which the transmit-to-receive coupling is the key parameter. As no low coupling configuration could be found a simple configuration is chosen and practical measurements are taken. These antennas are then used in the radar system that is to be built. An FM-CW radar system is designed and simulated using software with a negative feedback loop being designed and implemented into the radar simulation. A practical radar system is then made inclusive of the feedback loop. Measurements are then taken to determine the efficacy of the feedback loop. / AFRIKAANSE OPSOMMING: ʼn Bekende probleem met FM-CW radar stelsels is die lekkasie van versender krag tot in die ontvanger wat lei tot die maak van nabye teikens en kan die stelsel se dinamiese sendbereik steng beperk. Hierdie tesis oorweeg twee oplossings tot hierdie probleem vir ʼn lae frekwensie radar wat in die VHF band werk. Die eerste metode wat na gekyk word om die koppeling te onderdruk is om die twee antennas van die radar stelsel so te ontwerp sodat die hoeveelheid koppeling tussen hulle verminder is. Die tweede is om ʼn negatiewe terugvoerlus as deel van die ontvanger te ontwerp. Hierdie terugvoerlus sal die versender lekkasie sein aanpassend in die ontvanger verminder. In hierdie projek word die realisering van bogenoemde oplossings uiteengeset. ʼn Paar verskillende antenna ontwerpe word gemodelleer in sagteware en word gesimuleer om hul karakteristieke te bepaal. Die belangrikste van hierdie faktore is die versender na ontvanger koppeling. Sienend dat geen ontwerp met ʼn lae genoeg koppeling gevind kon word nie, is ʼn eenvoudige ontwerp gekies en praktiese metings daarvan geneem. Hierdie antennas word dan gebruik in die radar stelsel wat gebou sal word. ʼn FM-CW radar stelsel word ontwerp en gesimuleer in sagteware. Die negatiewe terugvoerlus word ook ontwerp en geïmplementeer in die radar simulasie. ʼn Praktiese radar stelsel word dan gemaak insluitend die terugvoerlus. Metings word dan geneem om die effektiwiteit daarvan te bepaal.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2006. / Speech is the primary and most natural means of communication between human beings. With the rapid spread of technology across the globe and the increased number of personal and public applications for digital equipment in recent years, the need for human/machine interaction has increased dramatically. Synthetic speech is audible speech produced by a machine automatically. A text-to-speech (TTS) system is one that converts bodies of text into digital speech signals which can be heard and understood by a person. Current TTS systems generally require large annotated speech corpora in the languages for which they are developed. For many languages these resources are not available. In their absence, a TTS system generates synthetic speech by means of mathematical algorithms constrained by certain rules. This thesis describes the design and implementation of a rule-based speech generation algorithm for use in a TTS system. The system allows the type, emphasis, pitch and other parameters associated with a sound and its particular mode of articulation to be specified. However, no attempt is made to model prosodic and other higher-level information. Instead, this is assumed known. The algorithm uses linear predictive (LP) models of monophone speech units, which greatly reduces the amount of data required for development in a new language. A novel approach to the interpolation of monophone speech units is presented to allow realistic transitions between monophone units. Additionally, novel algorithms for estimation and modelling of the harmonic and stochastic content of an excitation signal are presented. This is used to determine the amount of voiced and unvoiced energy present in individual speech sounds. Promising results were obtained when evaluating the developed system’s South African English speech output using two widely used speech intelligibility tests, namely the modified rhyme test (MRT) and semantically unpredictable sentences (SUS).

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: In recent years the demand for bandwidth has dramatically increased because of new applications for data and multimedia, and wireless technology has prevailed as a prominent technology for data connectivity, especially for home, office and last mile services. As wireless communications are dependant upon spectrum availability, which is communal, this scarce commodity in communication has to be used as efficiently as possible. Some aspects of this requirement are addressed in this project. We chose the IEEE 802.11b standard for this particular investigation because of its widespread use, the vast amount of applicable literature, the variety of software simulation tools and the ease with which equipment can be obtained. The IEEE 802.11 standard specified the Point Coordination Function as the de- terministic protocol. Recently research into this aspect has stagnated, and it was the purpose of this project to investigate how existing infrastructure networks could be improved by optimising some modes of the 802.11 protocol. The investigation also hoped to determine when to change between Distributed Coordination Function (DCF) and Point Coordination Function (PCF), and to provide an adaptive protocol to do so. This thesis presents mathematical models for the operation of DCF and PCF modes, which is compared with results from a network simulator (ns2), for theoretical veri- fication. A protocol is also proposed to dynamically switch between DCF and PCF, to harness the advantages they present. / AFRIKAANSE OPSOMMING: Die afgelope paar jaar het die aanvraag na bandwydte dramaties verhoog as gevolg van nuwe toepassings vir data en multimedia, en draadlose tegnologie het voorgekom as ’n dominante tegnologie vir data konnektiwiteit, veral vir die huis, kantoor en laaste myl dienste. Omdat draadlose kommunikasie afhanklik is van spektrum beskikbaarheid, wat gemeenskaplik is, moet hierdie skaars kommoditeit in kommunikasie so effektief moontlik gebruik word. Sekere aspekte van die vereiste sal in die tesis ondersoek word. Dit is besluit om die IEEE 802.11b standard vir die spesifieke ondersoek te gebruik as gevolg van die wye toepassing, die groot hoeveelheid beskikbare literatuur, die verskeidenheid simulasie sagteware en die gemak waarmee die toerusting bekom kan word. Die IEEE 802.11 standaard spesifiseer the Punt Koordinasie Funksie (PCF) as die deterministiese protokol vir die betrokke standaard. Onlangs het navorsing oor hierdie aspek gestagneer, en dit is die doel van die projek om te ondersoek hoe bestaande infrastruktuur netwerke moontlik verbeter kan word deur optimering van sekere modusse van die 802.11 protokol. Die ondersoek hoop ook om te bepaal wanneer die oorgang van die Distrubusie Koordinasie Funksie (DCF) en Punt Ko- ordinasie Funksie sal plaasvind, en om ’n dienooreenstemmende protokol te on- twikkel. Die tesis verskaf wiskundige modelle vir die werking van die DCF en PCF modusse, wat vergelyk word met resultate uit ’n netwerk simulator (ns2), vir teoretiese ver- ifikasie. ’n Protokol word ook voorgestel om dinamies te wissel tussen DCF and PCF, om die voordele wat die protokolle verskaf te gebruik.

Thesis (MScEng)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: The purpose of this study was to develop modi ed tract-etched membranes using nanocomposite TiO2 for advanced water treatment processes. Photocatalytic oxidation and reduction reactions take place on TiO2 surfaces under UV light irradiation, therefore sunlight and even normal indoor lighting could be utilised to achieve this effect. In membrane ltration, caking is a major problem, by enhancing the anti-fouling properties of photocatalysts to mineralise organic compounds the membrane life and e ciency can be improved upon. In this study the rst approach in nanocomposite membrane development was to directly modify the surface of polyethylenetherephthalate (PET) track-etched membranes (TMs) with titanium dioxide (TiO2) using inverted cylindrical magnetron sputtering (ICMS) for TiO2 thin lm deposition. The second approach was rst to thermally evaporate silver (Ag) over the entire TM surface, followed by sputtering TiO2 over the silver-coated TM. As a result a noble metal-titania nanocomposite thin lm layer is produced on top of the TM surface with both self-cleaning and superhydrophilic properties. Reactive inverted cylindrical magnetron sputtering is a physical vapour deposition method, where material is separated from a target using high energy ions and then re-assimilated on a substrate to grow thin lms. Argon gas is introduced simultaneously into the deposition chamber along with O2 (the reactive gas) to form TiO2. The photocatalytic activity and other lm properties, such as crystallinity can be in uenced by changing the sputtering power, chamber pressure, target-to-substrate distance, substrate temperature, sputtering gas composition and ow rate. These characteristics make sputtering the perfect tool for the preparation of di erent kinds of TiO2 lms and nanostructures for photocatalysis. In this work, the utilisation of ICMS to prepare photocatalytic TiO2 thin lms deposited on track-etched membranes was studied in detail with emphasis on bandgap reduction and TM surface regeneration. Nanostructured TiO2 photocatalysts were prepared through template directed deposition on track-etched membrane substrates by exploiting the good qualities of ICMS. The TiO2-TM as well as Ag-TiO2-TM thin lms were thoroughly characterised. ICMS prepared TiO2 lms were shown to exhibit good photocatalytic activities. However, the nanocomposite Ag-TiO2 thin lms were identi ed to be a much better choice than TiO2 thin lms on their own. Finally a clear enhancement in the photocatalytic activity was achieved by forming the Ag-TiO2 nanocomposite TMs. This was evident from the band-gap improvement from 3.05 eV of the TiO2 thin lms to the 2.76 eV of the Ag-TiO2 thin lms as well as the superior surface regenerative properties of the Ag-TiO2-TMs. / AFRIKAANSE OPSOMMING: Die doel van hierdie studie was om verbeterde baan-ge etste membrane (BMe) met behulp van nano-saamgestelde titaandioksied (TiO2) vir gevorderde water behandeling prosesse te ontwikkel. Fotokatalitiese oksidasie- en reduksie reaksies vind plaas op die TiO2 oppervlaktes onder UV-lig bestraling, en dus kan sonlig en selfs gewone binnenshuise beligting gebruik word om die gewenste uitwerking te verkry. In membraan ltrasie is die aanpaksel van onsuiwerhede 'n groot probleem, maar die verbetering van die self-reinigende eienskappe van fotokatalisators deur organiese verbindings te mineraliseer, kan die membraan se leeftyd en doeltre endheid verbeter word. In hierdie studie was die eerste benadering om nano-saamgestelde membraan ontwikkeling direk te verander deur die oppervlak van polyethylenetherephthalate (PET) BMe met 'n dun lagie TiO2 te bedek, met behulp van reaktiewe omgekeerde silindriese magnetron verstuiwing (OSMV).Die tweede benadering was eers om silwer (Ag) termies te verdamp oor die hele BM oppervlak, gevolg deur TiO2 verstuiwing bo-oor die silwer bedekte BM. As gevolg hiervan is 'n edelmetaal-titanium nano-saamgestelde dun lm laag gevorm bo-op die oppervlak van die BM, met beide self-reinigende en verhoogde hidro liese eienskappe. OSMV is 'n siese damp neerslag metode, waar materiaal van 'n teiken, met behulp van ho e-energie-ione, geskei word, en dan weer opgeneem word op 'n substraat om dun lms te vorm. Argon gas word gelyktydig in die neerslag kamer, saam met O2 (die reaktiewe gas), vrygestel om TiO2 te vorm. Die fotokatalitiese aktiwiteit en ander lm eienskappe, soos kristalliniteit, kan be nvloed word deur die verandering van byvoorbeeld die verstuiwingskrag, die druk in die reaksiekamer, teiken-tot-substraat afstand, substraattemperatuur, verstuiwing gassamestelling en vloeitempo. Hierdie eienskappe maak verstuiwing die ideale hulpmiddel vir die voorbereiding van die verskillende soorte TiO2 lms en nanostrukture vir fotokatalisasie. In hierdie tesis word OSMV gebruik ter voorbereiding van fotokatalitiese TiO2 dun lms, wat gedeponeer is op BMe. Hierdie lms word dan in diepte bestudeer, met die klem op bandgaping vermindering en BM oppervlak hergenerasie. Nanogestruktureerde TiO2 fotokataliste is voorberei deur middel van sjabloongerigte neerslag op BM substrate deur die ontginning van die goeie eienskappe van OSMV. Die TiO2-BM dun lms, sowel as Ag-TiO2-BM dun lms, is deeglik gekarakteriseer. OSMV voorbereide TiO2 dun lms toon goeie fotokatalitiese aktiwiteite. Nano-saamgestelde Ag-TiO2 dun lms is egter ge denti seer as 'n veel beter keuse as TiO2 dun lms. Ten slotte is 'n duidelike verbetering in die fotokatalitiese aktiwiteit bereik deur die vorming van die Ag-TiO2 nano-saamgestelde BMe. Dit was duidelik uit die bandgapingverbetering van 3,05 eV van TiO2 dun lms in vergelyking met die 2,76 eV van Ag-TiO2 dun lms. 'n Duidelike verbetering is behaal in die fotokatalitiese aktiwiteit deur die vorming van die Ag-TiO2 nano-saamgestelde TMs.