The emergence of a vast range of communication devices running on different types of technology has made convergence of technology become the order of the day. This revolution observed in communications technology has resulted in a pressing need for larger bandwidth, higher data rate and better spectrum availability, and it has become important that these factors be addressed. As such, this has resulted in the current resurgence of interest to investigate higher electromagnetic spectrum space that can take care of these needs. For the past decade, microwave (3 GHz-30 GHz) and millimeter waves (30 GHz-300 GHz) have been used as the appropriate frequency ranges for applications with properties such as wide bandwidth, smaller components size, narrow beamwidths, frequency re-use, small antenna, and short deployment time. To optimize the use of these frequency ranges by communication systems, the three tiers of communication system elements - receiver, transmitter and transmission channel or medium must be properly designed and configured. However, if the transmitter and receiver meet the necessary requirements, the medium in which signals are transmitted often becomes an issue at this range of frequencies. The most significant factor that affects the transmission of signals at these bands is attenuation and scattering by rain, snow, water vapour and other gases in the atmosphere. Scattering and absorption by rain at microwave and millimeter bands is thus a main concern for system designers. This study presents results of research into the interaction of rainfall with microwave and millimeter wave propagation as a medium. The study of rainfall characteristics allows estimation of its scattered and attenuated effects in the presence of microwave and millimeter waves. The components of this work encompass rainfall rate integration time, cumulative distribution and modelling of rainfall rate and characteristics of rain drop size and its modelling. The effects of rain on microwave and millimeter wave signals, which result in rain attenuation, are based on rainfall rate variables such as rainfall rate cumulative distribution, raindrop size distribution, total scattering cross sections, rain drop shape, and rain drop terminal velocity. A regional rainfall rate conversion factor from five-minute rainfall data to one-minute integration time is developed using the existing conversion method and a newly developed hybrid method. Based on these conversion factor results from the hybrid method, the rainfall at five-minute integration time was converted to a one-minute equivalent to estimate its cumulative distributions. In addition, new rain zones based on ITU-R and Crane designations are suggested for the entire region of South Africa and the surrounding Islands. The results are compared with past research work done in the other regions. Rain attenuation is acutely influenced by rain drop size distribution (DSD). This study thus also investigates DSD models from previous research work. There are several DSD models commonly used to estimate rain attenuation. They are models which have their root from exponential, gamma, lognormal and Weibull distributions. Since DSD is dynamic and locationdependent, a simple raindrop size distribution model is developed for Durban using maximum likelihood estimation (MLE) method. The MLE method is applied to the three-parameter lognormal distribution in order to model DSD for Durban. Rain drop size depends on rainfall rate, drop diameter and rain drop velocity. Semi-empirical models of terminal velocity from previous studies are investigated in this work and proposed for the estimation of specific rain attenuation. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2010.

Telecommunications in the current information age is increasingly relying on the wireless link. This is because wireless communication has made possible a variety of services ranging from voice to data and now to multimedia. Consequently, demand for new wireless capacity is growing rapidly at a very alarming rate. In a bid to cope with challenges of increasing demand for higher data rate, better quality of service, and higher network capacity, there is a migration from Single Input Single Output (SISO) antenna technology to a more promising Multiple Input Multiple Output (MIMO) antenna technology. On the other hand, Orthogonal Frequency Division Multiplexing (OFDM) technique has emerged as a very popular multi-carrier modulation technique to combat the problems associated with physical properties of the wireless channels such as multipath fading, dispersion, and interference. The combination of MIMO technology with OFDM techniques, known as MIMO-OFDM Systems, is considered as a promising solution to enhance the data rate of future broadband wireless communication Systems. This thesis addresses a major area of challenge to both SISO-OFDM and MIMO-OFDM Systems; estimation of accurate channel state information (CSI) in order to make possible coherent detection of the transmitted signal at the receiver end of the system. Hence, the first novel contribution of this thesis is the development of a low complexity adaptive algorithm that is robust against both slow and fast fading channel scenarios, in comparison with other algorithms employed in literature, to implement soft iterative channel estimator for turbo equalizer-based receiver for single antenna communication Systems. Subsequently, a Fast Data Projection Method (FDPM) subspace tracking algorithm is adapted to derive Channel Impulse Response Estimator for implementation of Decision Directed Channel Estimation (DDCE) for Single Input Single Output - Orthogonal Frequency Division Multiplexing (SISO-OFDM) Systems. This is implemented in the context of a more realistic Fractionally Spaced-Channel Impulse Response (FS-CIR) channel model, as against the channel characterized by a Sample Spaced-Channel Impulse Response (SS)-CIR widely assumed by other authors. In addition, a fast convergence Variable Step Size Normalized Least Mean Square (VSSNLMS)-based predictor, with low computational complexity in comparison with others in literatures, is derived for the implementation of the CIR predictor module of the DDCE scheme. A novel iterative receiver structure for the FDPM-based Decision Directed Channel Estimation scheme is also designed for SISO-OFDM Systems. The iterative idea is based on Turbo iterative principle. It is shown that improvement in the performance can be achieved with the iterative DDCE scheme for OFDM system in comparison with the non iterative scheme. Lastly, an iterative receiver structure for FDPM-based DDCE scheme earlier designed for SISO OFDM is extended to MIMO-OFDM Systems. In addition, Variable Step Size Normalized Least Mean Square (VSSNLMS)-based channel transfer function estimator is derived in the context of MIMO Channel for the implementation of the CTF estimator module of the iterative Decision Directed Channel Estimation scheme for MIMO-OFDM Systems in place of linear minimum mean square error (MMSE) criterion. The VSSNLMS-based channel transfer function estimator is found to show improved MSE performance of about -4 MSE (dB) at SNR of 5dB in comparison with linear MMSE-based channel transfer function estimator. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2010.

Advances in today’s fast growing communication systems have resulted in congestion in the lower frequency bands and the need for higher capacity broadband services. This has made it inevitable for service providers to migrate to higher frequency bands so as to accommodate the ever increasing demands on radio communication systems. However, the reliability of such systems at these frequency bands tend to be severely degraded due to some natural atmospheric phenomena of which rain is the dominant factor. This is not to say that other factors have become unimportant; however, if attenuation by rain is so severe that a radio link is unavailable for use, then other factors become secondary. Therefore, it is paramount to establish a model capable of predicting the behaviour of these systems in the presence of rain. This study employs a semi-empirical approach for the formulation of rain attenuation models using the knowledge of rain rate, raindrop size distribution, and a signal level measurement recorded at 19.5 GHz on a horizontally polarized terrestrial radio link. The semi-empirical approach was developed by considering the scattering effect of an electromagnetic wave propagating through a medium containing raindrops. The complex forward scattering amplitudes for the raindrops are determined for all raindrop sizes at different frequencies, utilizing the Mie scattering theory on spherical dielectric raindrops. From these scattering amplitudes, the extinction cross-sections for the spherical raindrops are calculated. Applying the power-law regression to the real part of the calculated extinction cross-section, power-law coefficients are determined at different frequencies. The power-law model generated from the extinction crosssection is integrated over different raindrop-size distribution models to formulate theoretical rain attenuation models. The developed rain attenuation models are used with 0.01 R rain rate statistics determined for four locations in different rain climatic zones in South Africa to calculate the specific rain attenuation. From a horizontally polarized 6.73 km terrestrial line-of-sight link in Durban, South Africa,experimental rain attenuation measurements were recorded at 19.5 GHz. These rain attenuation measurements are compared with the results obtained from the developed attenuation models with the same propagation parameters to establish the most appropriate attenuation models that describe the behaviour of radio link performance in the presence of rain. For the purpose of validating the results, it is compared with the ITU-R rain attenuation model. This study also considers the characteristics and variations associated with rain attenuation for terrestrial communication systems. This is achieved by utilizing the ITU-R power-law rain attenuation model on 5-year rain rate data obtained from the four different climatic rain zones in South Africa to estimate the cumulative distributions of rain attenuation. From the raindrop size and 1-minute rain rate measurement recorded in Durban with a distrometer over six months, rain events over the six months are classified into drizzle, widespread, shower and thunderstorm rain types and the mean rain rate statistics determined for each class of rain. Drop-size distribution for all the rain types is estimated. This research has shown a statistical analysis of rain fade data and proposed an empirical rain attenuation model for South Africa localities. This work has also drawn out theoretical rain attenuation prediction models based on the assumption that the shapes of raindrops are spherical. The results predicted from these theoretical attenuation models have shown that it is not the raindrop shapes that determine the attenuation due to rain, but the raindrop size distribution and the rain rate content in the drops. This thesis also provides a good interpretation of cumulative rain attenuation distribution on seasonal and monthly basis. From these distributions, appropriate figures of fade margin are derived for various percentages of link availability in South Africa. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2010.

This thesis has investigated radioclimatological study in a clear-air environment as applicable to terrestrial line of sight link design problems. Radioclimatological phenomena are adequately reviewed both for the precipitation effect and clear-air effect. The research focuses more on the clear-air effect of radioclimatological studies. Two Southern African countries chosen for case study in the report are Botswana and South Africa. To this end, radiosonde data gathered in Maun, Botswana and Durban, South Africa are used for model formulation and verification. The data used in the thesis ranges from three years to ten years in these two stations. Three to ten years of refractivity data gathered in Botswana and South Africa is used for the model formulation. On the other hand, eight months signal level measurement data recorded from the terrestrial line of sight link set up between Howard College and Westville Campuses of the University of KwaZulu-Natal, Durban South Africa is used for model verification. Though various radioclimatic parameters could affect radio signal propagation in the clear-air environment, this report focuses on two of these parameters. These two parameters are the geoclimatic factor and effective earth radius factor (k-factor). The first parameter is useful for multipath fading determination while the second parameter is very important for diffraction fading, modeling and characterization. The two countries chosen have different terrain and topographical structures; thus further underlying the choice for these two parameters. While Maun in Botswana is a gentle flat terrain, Durban in South Africa is characterized by hilly and mountainous terrain structure, which thus affects radioclimatological modeling in the two countries. Two analytical models have been proposed to solve clear-air radioclimatic problems in Southern Africa in the thesis. The first model is the fourth order polynomial analytical expression while the second model is the parabolic equation. The fourth order polynomial model was proposed after an extensive analysis of the eight month signal level measurement data gathered in Durban, South Africa. This model is able to predict the fade exceedance probabilities as a function of fade depth level. The result from the fourth order polynomial model is found to be comparable with other established multipath propagation model reviewed in the thesis. Availability of more measurement data in more location will be necessary in future to further refine this model. The second model proposed to solve clear-air propagation problem in the thesis is the modified parabolic equation. We chose this technique because of its strength and its simplistic adaptation to terrestrial line of sight link design problem. This adaptation is possible because, the parabolic equation can be modified to incorporate clear-air parameters. Hence this modification of the parabolic equation allows the possibility of a hybrid technique that incorporates both the statistical and mathematical procedures perfectly into one single process. As a result of this, most of the very important phenomena in clear-air propagation such as duct occurrence probabilities, diffraction fading and multipath fading is captured by this technique. The standard parabolic equation (SPE) is the unmodified parabolic equation which only accounts for free space propagation, while the modified parabolic equation (MPE) is the modified version of the parabolic equation. The MPE is classified into two in the thesis: the first modified parabolic equation (MPE1) and second modified parabolic equation (MPE2). The MPE1 is designed to incorporate the geoclimatic factor which is intended to study the multipath fading effect in the location of study. On the other hand, MPE2 is the modified parabolic equation designed to incorporate the effective earth radius factor (k-factor) intended to study the diffraction fading in the location of study. The results and analysis of the results after these modifications confirm our expectation. This result shows that signal loss is due primarily to diffraction fading in Durban while in Botswana, signal loss is due primarily to multipath. This confirms our expectation since a flatter terrain attracts signal loss due to multipath while hilly terrain attracts signal loss due to diffraction fading. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2010.

The last six years has seen the emergence and rapid development of a new type of field effect transistor, the High Electron Mobility Transistor (HEMT), which offers improved performance in both digital and analogue circuits compared with circuits incorporating either MEtal Semiconductor (MES) or Metal Oxide Semiconductor (MOS) FETs. A new physically-based analytic model for HEMTs, which predicts the DC and RF electrical performance from the material and structural parameters of the device, is presented. The efficacy of the model is demonstrated with comparisons between simulated and measured device characteristics, at DC and microwave frequencies. The good agreement with experiment obtained with the model indicates that velocity overshoot effects are considerably less important in HEMTs than has been widely assumed, and that the electron transit velocity in submicron devices is approximately 10 cm/s, rather than around 2x10 cm/s. The Inverted HEMT, one of the major HEMT structural variants, is emphasized throughout this work because of its potential advantages over other variants, and practical results from 0.5 micron gate length Inverted HEMTs are presented. / Thesis (Ph.D.)-University of Natal, Durban, 1986.

Travelling-wave distributed amplifiers are providing gain over broad frequency ranges for microwave applications. Similar concepts are applicable to distributed mixers and, with the use of controlled feedback, to a multifunction component simultaneously emulating a mixer, amplifier and an oscillator. The concept of this new travelling-wave frequency converter is introduced and data for a discrete component test circuit is presented. To facilitate the converter operation a new three-port travelling-wave mixer is introduced and characterized. Four-port scattering and wave scattering transformations are derived as a method of analysis of the four-port distributed structure. This enables sequential circuit analysis on a small computer. Practical applications unique to the advanced automatic network analyser, including time domain measurements, are presented to characterize test circuits as well as to develop ancillary equipment such as a transistor test fixture. Automated error corrected transistor measurements and de-embedding are also discussed. A piecewise linear quantum mechanical method of modelling the conduction channel of a short gate field effect transistor is given to aid the extrapolation of the distributed frequency converter concept to submicron and heterojunction structures. / Thesis (Ph.D.)-University of Natal, Durban, 1985.

Providing guarantees in the Quality of Service (QoS) has become essential to the transmission of multimedia traffic over wireless links with fading channels. However this poses significant challenges due to the variable nature of such channels and the diverse QoS requirements of different applications including voice, video and data. The benefits of dynamic adaptation to system and channel conditions have been accepted, but the true potential of optimized adaptation is lost if the layers operate independently, ignoring possible interdependencies between them. Cross-layer design mechanisms exploit such interdependencies to provide QoS guarantees for the transmission of multimedia traffic over fading channels. Channel adaptive M-QAM schemes are examples of some of the earliest works in the area of cross-layer design. However, many of the original schemes use the assumption that thresholds designed for AWGN channels can be directly applied to slow-fading channels. The thresholds are calculated with a commonly used approximation bit error rate (BER) expression and the first objective of the thesis was to study the accuracy of this commonly used expression in fading channels. It is shown that that the inaccuracy of the expression makes it unsuitable for use in the calculation of the threshold points for an adaptive M-QAM system over fading channels. An alternative BER expression is then derived which is shown to be far more accurate than the previous one. The improved accuracy is verified through simulations of the system over Nakagami-m fading channels. Many of the cross-layer adaptation mechanisms that address the QoS provisioning problem only use the lower layers (physical and data link) and few explore the possibility of using higher layers. As a result, restrictions are placed on the system which introduces functional limitations such as the inability to insert more than one class of traffic in a physical layer frame. The second objective in this thesis was to design a physical and application layer cross-layer adaptation mechanism which overcomes this limitation. The performance results of the scheme in both AWGN and fading channels show that the cross-layer mechanism can be efficiently utilized for the purposes of providing error rate QoS guarantees for multimedia traffic transmissions over wireless links. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2009.

Though declining in popularity over the last decade, the magnetron still has applications where portable high power is needed. This study examines the predicted performance of cylindrical microwave magnetrons using analytic lumped-spoke models based on the energy conservation principle. The analytic approach is still favoured when small computer systems are used and the overall performance of the tube is to be predicted. The magnetron elements are examined and the role they play in the overall device performance analysed. Simplified representations of these elements are used to construct a complete magnetron model. The Hartree threshold condition is reexamined and a new, more accurate analytic formulation proposed. This formulation is based on electric field strengths at the base of the magnetron spoke. The effect of the space charge on the threshold condition is included. Spoke current has been evaluated at the edge of the Brillouin hub. The resulting anode- cathode voltage performance predictions are consistent with measured results. A computer program has been written to analyse the performance predicted by this model. Models proposed by other authors are examined, and compared to this model. The resulting model has been tested by comparing predicted results to the measured performance of four slot-and-hole magnetrons. To facilitate accurate magnetron testing, a new automated triple-stub high power microwave load has been developed. The load operates at a peak power of IMW from 2,7-3,OGHZ, and allows the change of the VSWR to any value along any path within the VSWR=I, 5: I circle. The development of the triple tuner and termination is discussed in detail. A new waterload configuration which has the advantage of simple construction yet good matching characteristics is presented. Automated measurement of pulling figure and construction of Rieke diagrams is discussed. The accuracy of the complete load is compared to conventional loads currently in service in the tube industry. / Thesis (Ph.D.)-University of Natal, Durban, 1988.

Investigations into currently accepted methods of modelling variations in the arrival rate of meteors over forward-scatter meteor links have revealed some shortcomings. In these investigations, particular emphasis is placed on the work of Rudie due to its current acceptance in the literature. The non-uniform radiant distribution of meteors measured by Davies and modelled by Rudie, is critically examined and predictions using these models are compared with measured results taken over two forward-scatter links in the Southern Hemisphere. A new, alternative method of including the effect of non-uniform radiant distributions on the predicted arrival rate of meteors is given. The method used by Rudie to model Davies' measured radiant distribution is shown to be unnecessarily complicated and a simpler alternative is given. Furthermore, Rudie's distribution is shown not to be derived from a particular set of Davies' results as implied by Rudie. Other non-uniform distributions of meteors are also investigated. Comparisons between monthly-averaged daily cycles of measured and predicted arrival rates of meteors for a midpath and an endpath meteor link are used to reveal the validity and limitations of the published distributions. A new graphical method is used to aid in determining the validity and limitations of the non-uniform distributions. Discrepancies in the published predicted and measured annual variations in the arrival rate of meteors are investigated. Contrary to recently published information, predicted annual variations in the arrival rate of meteors for meteor radiants close to the ecliptic are shown to be comparable to measured results. / Thesis (Ph.D.)-University of Natal, Durban, 1991.

The ever increasing demand for fast and efficient broadband wireless communication services requires future broadband communication systems to provide a high data rate, robust performance and low complexity within the limited available electromagnetic spectrum. One of the identified, most-promising techniques to support high performance and high data rate communication for future wireless broadband services is the deployment of multi-input multi-output (MIMO) antenna systems with orthogonal frequency division multiplexing (OFDM). The combination of MIMO and OFDM techniques guarantees a much more reliable and robust transmission over a hostile wireless channel through coding over the space, time and frequency domains. In this thesis, two full-rate space-time coded OFDM systems are proposed. The first one, designed for two transmit antennas, is called extended super-orthogonal space-time trellis coded OFDM (ESOSTTC-OFDM), and is based on constellation rotation. The second one, called super-quasi-orthogonal space-time trellis coded OFDM (SQOSTTCOFDM), combines a quasi-orthogonal space-time block code with a trellis code to provide a full-rate code for four transmit antennas. The designed space-time coded MIMO-OFDM systems achieve a high diversity order with high coding gain by exploiting the diversity advantage of frequency-selective fading channels. Concatenated codes have been shown to be an effective technique of achieving reliable communication close to the Shannon limit, provided that there is sufficient available diversity. In a bid to improve the performance of the super orthogonal space-time trellis code (SOSTTC) in frequency selective fading channels, five distinct concatenated codes are proposed for MIMO-OFDM over frequency-selective fading channels in the second part of this thesis. Four of the coding schemes are based on the concatenation of convolutional coding, interleaving, and space-time coding, along multiple-transmitter diversity systems, while the fifth coding scheme is based on the concatenation of two space-time codes and interleaving. The proposed concatenated Super-Orthogonal Space-Time Turbo-Coded OFDM System I. B. Oluwafemi 2012 vii coding schemes in MIMO-OFDM systems achieve high diversity gain by exploiting available diversity resources of frequency-selective fading channels and achieve a high coding gain through concatenations by employing the turbo principle. Using computer software simulations, the performance of the concatenated SOSTTC-OFDM schemes is compared with those of concatenated space-time trellis codes and those of conventional SOSTTC-OFDM schemes in frequency-selective fading channels. Simulation results show that the concatenated SOSTTC-OFDM system outperformed the concatenated space-time trellis codes and the conventional SOSTTC-OFDM system under the various channel scenarios in terms of both diversity order and coding gain. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2012.