Multiple-input multiple-output (MIMO) systems have been extensively studied in the past decade. The attractiveness of MIMO systems is due to the fact that they drastically reduce the deleterious e ects of multipath fading leading to high system capacity and low error rates. In situations where wireless devices are restrained by their size and hardware complexity, such as mobile phones, transmit diversity is not achievable. A new paradigm called cooperative communication is a viable solution. In a cooperative scenario, a single-antenna device is assisted by another single-antenna device to relay its message to the destination or base station. This creates a virtual multiple-input multiple-output (MIMO) system. There exist two cooperative strategies: amplify-and-forward (AF) and decode-and-forward (DF). In the former, the relay ampli es the noisy signal received from the source before forwarding it to the destination. No form of demodulation is required. In the latter, the relay rst decodes the source signal before transmitting an estimate to the destination. In this work, focus is on the DF method. A drawback of an uncoded DF cooperative strategy is error propagation at the relay. To avoid error propagation in DF, various relay selection schemes can be used. Coded cooperation can also be used to avoid error propagation at the relay. Various error correcting codes such as convolutional codes or turbo codes can be used in a cooperative scenario. The rst part of this work studies a variation of the turbo codes in cooperative diversity, that further reduces error propagation at the relay, hence lowering the end-to-end error rate. The union bounds on the bit-error rate (BER) of the proposed scheme are derived using the pairwise error probability via the transfer bounds and limit-before-average techniques. In addition, the outage analysis of the proposed scheme is presented. Simulation results of the bit error and outage probabilities are presented to corroborate the analytical work. In the case of outage probability, the computer simulation results are in good agreement with the the analytical framework presented in this chapter. Recently, most studies have focused on cross-layer design of cooperative diversity at the physical layer and truncated automatic-repeat request (ARQ) at the data-link layer using the system throughput as the performance metric. Various throughput optimization strategies have been investigated. In this work, a cross-relay selection approach that maximizes the system throughput is presented. The cooperative network is comprised of a set of relays and the reliable relay(s) that maximize the throughput at the data-link layer are selected to assist the source. It can be shown through simulation that this novel scheme outperforms from a throughput point of view, a system throughput where the all the reliable relays always participate in forwarding the source packet. A power optimization of the best relay uncoded DF cooperative diversity is investigated. This optimization aims at maximizing the system throughput. Because of the non-concavity and non-convexity of the throughput expression, it is intractable to derive a closed-form expression of the optimal power through the system throughput. However, this can be done via the symbol-error rate (SER) optimization, since it is shown that minimizing the SER of the cooperative system is equivalent to maximizing the system throughput. The SER of the retransmission scheme at high signal-to-noise ratio (SNR) was obtained and it was noted that the derived SER is in perfect agreement with the simulated SER at high SNR. Moreover, the optimal power allocation obtained under a general optimization problem, yields a throughput performance that is superior to non-optimized power values from moderate to high SNRs. The last part of the work considers the throughput maximization of the multi-relay adaptive DF over independent and non-identically distributed (i.n.i.d.) Rayleigh fading channels, that integrates ARQ at the link layer. The aim of this chapter is to maximize the system throughput via power optimization and it is shown that this can be done by minimizing the SER of the retransmission. Firstly, the closed-form expressions for the exact SER of the multi-relay adaptive DF are derived as well as their corresponding asymptotic bounds. Results showed that the optimal power distribution yields maximum throughput. Furthermore, the power allocated at a relay is greatly dependent of its location relative to the source and destination. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2013.

This thesis investigates the use of correlation techniques to perform system identification tests, with the objective of developing online test methods to perform mechanical parameter extraction as well as machine diagnostics. More specifically, these test methods must be implemented on a Programmable Logic Controller (PLC) in combination with Variable Speed Drives (VSD). Models for motor-based mechanical systems are derived and other documented methods for parameter identification of mechanical systems are discussed. An investigation is undertaken into the principle that the impulse response of a system may be obtained when a test signal with an impulsive autocorrelation is injected into the system. The theory of using correlation functions to determine the numerical impulse response of a system is presented. Suitable test signals, pseudorandom binary sequences (PRBS) are analysed, and their generation and properties are discussed. Simulations are presented as to how the various properties of the PRBS test signals influence the resulting impulse response curve. Further simulations are presented that demonstrate how PRBS-based tests in conjunction with a curve-fitting method, in this case the method of linear least squares, can provide a fair estimation of the parameters of a mechanical system. The implementation of a correlation based online testing routine on a PLC is presented. Results from these tests are reviewed and discussed. A SCADA system that has been designed is discussed and it is shown how this system allows the user to perform diagnostics on networked drives in a distributed automation system. Identification of other mechanical phenomena such as elasticity and the non-linearity introduced by the presence of backlash is also investigated. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2009.

Over the last decade, access to personal wireless communication networks has evolved to a point of necessity. Attached to the phenomenal growth of the telecommunications industry in recent times is an escalating demand for higher data rates and efficient spectrum utilization. This demand is fuelling the advancement of third generation (3G), as well as future, wireless networks. Current 3G technologies are adding a dimension of mobility to services that have become an integral part of modem everyday life. Wideband code division multiple access (WCDMA) is the standardized multiple access scheme for 3G Universal Mobile Telecommunication System (UMTS). As an air interface solution, CDMA has received considerable interest over the past two decades and a great deal of current research is concerned with improving the application of CDMA in 3G systems. A factoring component of CDMA is multiuser detection (MUD), which is aimed at enhancing system capacity and performance, by optimally demodulating multiple interfering signals that overlap in time and frequency. This is a major research problem in multipoint-to-point communications. Due to the complexity associated with optimal maximum likelihood detection, many different sub-optimal solutions have been proposed. This focus of this dissertation is the application of neural networks for MUD, in a direct sequence CDMA (DS-CDMA) system. Specifically, it explores how the Hopfield recurrent neural network (RNN) can be employed to give yet another suboptimal solution to the optimization problem of MUD. There is great scope for neural networks in fields encompassing communications. This is primarily attributed to their non-linearity, adaptivity and key function as data classifiers. In the context of optimum multiuser detection, neural networks have been successfully employed to solve similar combinatorial optimization problems. The concepts of CDMA and MUD are discussed. The use of a vector-valued transmission model for DS-CDMA is illustrated, and common linear sub-optimal MUD schemes, as well as the maximum likelihood criterion, are reviewed. The performance of these sub-optimal MUD schemes is demonstrated. The Hopfield neural network (HNN) for combinatorial optimization is discussed. Basic concepts and techniques related to the field of statistical mechanics are introduced and it is shown how they may be employed to analyze neural classification. Stochastic techniques are considered in the context of improving the performance of the HNN. A neural-based receiver, which employs a stochastic HNN and a simulated annealing technique, is proposed. Its performance is analyzed in a communication channel that is affected by additive white Gaussian noise (AWGN) by way of simulation. The performance of the proposed scheme is compared to that of the single-user matched filter, linear decorrelating and minimum mean-square error detectors, as well as the classical HNN and the stochastic Hopfield network (SHN) detectors. Concluding, the feasibility of neural networks (in this case the HNN) for MUD in a DS-CDMA system is explored by quantifying the relative performance of the proposed model using simulation results and in view of implementation issues. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, 2006.

In today's rapidly expanding communications industry, there is an ever-increasing demand for greater bandwidth, higher data rates and better spectral efficiency. As a result current and future communication systems will need to employ advanced spatial, temporal and frequency diversity techniques in order to meet these demands. Even with the utilisation of such techniques, the congestion of the lower frequency bands, will inevitably lead to the increased usage of the millimetre-wave frequencies in terrestrial communication systems. Before such systems can be deployed, radio system designers require realistic and readily useable channel and propagation models at their disposal to predict the behaviour of such communication links and ensure that reliable and efficient data transmission is achieved The scattering and attenuation of electromagnetic waves by rain is a serious problem for microwave and millimetre-wave frequencies. The conversion of rain rate to specific attenuation is a crucial step in the analysis of the total path attenuation and hence radio-link availability. It is now common practice to relate the specific attenuation and the rain rate using the simple power law relationship. The power-law parameters are then used in the path attenuation model, where the spatial variations of rainfall are estimated by a path-integration of the rain rate. These power law parameters are strongly influenced by the drop-size-distribution (DSD). Thus an examination of the various DSDs and their influence on the specific attenuation and link availability is warranted. Several models for the DSD have been suggested in literature, from the traditional exponential, to the gamma, log normal and Weibull distributions. The type of DSD varies depending on the geographical location and rainfall type. An important requirement of the DSD is that it is consistent with rain rate (i.e. the DSD must satisfy the rain-rate integral equation). Thus before application in the specific attenuation calculations, normalisation needs to be performed to ensure the consistency, as done in this study. Once the specific attenuation has been evaluated for necessary frequency and rain-rate range, path averaging is performed to predict the rain attenuation over the communication link. The final step in this dissertation is the estimation of the percentage of time of such occurrences. For this, cumulative time statistics of surface point rain rates are needed. The resulting cumulative distribution model of the fade depth and duration due to rain is a valuable tool for system designers. With such models the system designer can then determine the appropriate fade margin for the communication system and resulting period of unavailability for the link / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2006.

Mobile ad hoc networks (MANETs) eliminate the need for pre-existing infrastructure by relying on the nodes to perform all network services. The connectivity between the nodes is sporadic due to the shared, error-prone wireless medium and frequent route failures caused by node mobility. Fully self-organized MANETs are created solely by the end-users for a common purpose in an ad hoc fashion. Forming peer-to-peer security associations in MANETs is more challenging than in conventional networks due to the lack of central authority. This thesis is mainly concerned with peer- t o-peer key management in fully self-organized M ANETs. A key management protocol’s primary function is to bootstrap and maintain the security associations in the network, hence to create, distribute and revocate (symmetric or asymmetric) keying material as needed by the network security services. The fully self-organized feature means that t he key management protocol cannot rely on any form of off-line or on-line trusted third party (TTP). The first part of the thesis gives an introduction to MANETs and highlights MANETs' main characteristics and applications. The thesis follows with an overall perspective on the security issues in MANETs and motivates the importance of solving the key management problem in MANETs. The second part gives a comprehensive survey on the existing key management protocols in MANETs. The protocols are subdivided into groups based on their main characteristic or design strategy. Discussion and comments are provided on the strategy of each group. The discussions give insight into the state of the art and show researchers the way forward. The third part of the thesis proposes a novel peer- to-peer key management scheme for fully self-organized MANETs, called Self-Organized Peer-to-Peer Key Management (SelfOrgPKM). The scheme has low implementation complexity and provides self-organized mechanisms for certificate dissemination and key renewal without the need for any form of off-line or on-line authority. The fully distributed scheme is superior in communication and computational overhead with respect to its counterparts. All nodes send and receive the same number of messages and complete the same amount of computation. ScifOrgPKM therefore preserves the symmetric relationship between the nodes. Each node is its own authority domain which provides an adversary with no convenient point of attack. SelfOrgPKM solves t he classical routing-security interdependency problem and mitigates impersonation attacks by providing a strong one-to-one binding between a user’s certificate information and public key. The proposed scheme uses a novel certificate exchange mechanism t hat exploits user mobility but does not rely on mobility in anyway. The proposed certificate exchange mechanism is ideally suited for bootstraping the routing security. It enables nodes to setup security associations on the network layer in a localized fashion without any noticeable time delay. The thesis also introduces two generic cryptographic building blocks as the basis of SelfOrgPKM: 1) A variant on the ElGamal type signature scheme developed from the generalized ElGamal signature scheme introduced by Horster et al. The modified scheme is one of the most efficient ElGamal variants, outperforming most other variant s; and 2) A subordinate public key generation scheme. The thesis introduces t he novel notion of subordinate public keys, which allows the users of SelfOrgPKM to perform self-organized, self-certificate revocation without changing their network identifiers / addresses. Subordinate public keys therefore eliminate the main weakness of previous efforts to solve the address ownership problem in Mobile IPv6. Furthermore, the main weakness of previous efforts to break t he routing-security interdependence cycle in MANETs is also eliminated by a subordinate public key mechanism. The presented EIGamal signature variant is proved secure in t he Random Oracle and Generic Security Model (ROM+ GM ) without making any unrealistic assumptions . It is shown how the strong security of the signature scheme supports t he security of t he proposed subordinate key generation scheme. Based on the secure signature scheme a security argument for SelfOrgPKM is provided with respect to a genera l, active insider adversary model. The only operation of SelfOrgPKM affecting the network is the pairwise exchange of certificates. The cryptographic correctness, low implementation complexity and effectiveness of SelfOrgPKM were verified though extensive simulations using ns-2 and OpenSSL. Thorough analysis of the simulation results shows t hat t he localized certificate exchange mechanism on the network layer has negligible impact on network performance. The simulation results also correlate with efficiency analysis of SelfOrgPKM in an ideal network setting, hence assuming guaranteed connectivity. The simulation results furthermore demonstrate that network layer certificate exchanges can be triggered without extending routing protocol control packet. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2005.

The ability to track articulated human motion in video sequences is essential for applications ranging from biometrics, virtual reality, human-computer interfaces and surveillance. The work presented in this thesis focuses on tracking and analysing human motion in terms of MPEG-4 Body Animation Parameters, in the context of a model-based coding scheme. Model-based coding has emerged as a potential technique for very low bit-rate video compression. This study emphasises motion reconstruction rather than photorealistic human body modelling, consequently a 3-D skeleton with 31 degrees-of-freedom was used to model the human body. Compression is achieved by analysing the input images in terms of the known 3-D model and extracting parameters that describe the relative pose of each segment. These parameters are transmitted to the decoder which synthesises the output by transforming the default model into the correct posture. The problem comprises two main aspects: 3-D human motion capture and pose description. The goal of the 3-D human motion capture component is to generate 3-D locations of key joints on the human body without the use of special markers or sensors placed on the subject. The input sequence is acquired by three synchronised and calibrated CCD cameras. Digital image matching techniques including cross-correlation and least squares matching are used to find spatial correspondences between the multiple views as well as temporal correspondences in subsequent frames with sub-pixel accuracy. The tracking algorithm automates the matching process examining each matching result and adaptively modifying matching parameters. Key points must be manually selected in the first frame, following which the tracking commences without the intervention of the user, employing the recovered 3-D motion of the skeleton model for prediction of future states. Epipolar geometry is exploited to verify spatial correspondences in each frame before the 3-D locations of all joints are computed through triangulation to construct the 3-D skeleton. The pose of the skeleton is described by the MPEG-4 Body Animation Parameters. The subject's motion is reconstructed by applying the animation parameters to a simplified version of the default MPEG-4 skeleton. The tracking algorithm may be adapted to 2-D tracking in monocular sequences. An example of 2-D tracking of facial expressions demonstrates the flexibility of the algorithm. Further results involving tracking separate body parts demonstrate the advantage of multiple views and the benefit of camera calibration, which simplifies the generation of 3-D trajectories and the estimation of epipolar geometry. The overall system is tested on a walking sequence where full body motion capture is performed and all 31 degrees-of freedom of the tracked model are extracted. Results show adequate motion reconstruction (i.e. convincing to most human observers), with slight deviations due to lack of knowledge of the volumetric property of the human body. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2005.

Future wireless communication systems promise to offer a variety of multimedia services which require reliable transmission at high data rates over wireless links. Multiple input multiple output (MIMO) systems have received a great deal of attention because they provide very high data rates for such links. Theoretical studies have shown that the quality provided by MIMO systems can be increased by using space-time codes. Space-time codes combine both space (antenna) and time diversity in the transmitter to increase the efficiency of MIMO system. The three primary approaches, layered spacetime architecture, space-time trellis coding (STTC) and space-time block coding (STBC) represent a way to investigate transmitter-based signal processing for diversity exploitation and interference suppression. The advantages of STBC (i.e. low decoding complexity) and STTC (i.e. TCM encoder structure) can be used to design a high rate space-time trellis coded modulation (HR-STTCM). Most space-time codes designs are based on the assumption of perfect channel state information at the receiver so as to make coherent decoding possible. However, accurate channel estimation requires a long training sequence that lowers spectral efficiency. Part of this dissertation focuses on the performance of HR-STTCM under non-coherent detection where there is imperfect channel state information and also in environment where the channel experiences rapid fading. Prior work on space-time codes with particular reference to STBC systems in multiuser environment has not adequately addressed the performance of the decoupled user signalto-noise ratio. Part of this thesis enumerates from a signal-to-noise ratio point of view the performance of the STBC systems in multiuser environment and also the performance of the HR-STTCM in such environment. The bit/frame error performance of space-time codes in fading channels can be evaluated using different approaches. The Chemoff upper-bound combined with the pair state generalized transfer function bound approach or the modified state transition diagram transfer function bound approach has been widely used in literature. However, although readily detennined, this bound can be too loose over nonnal signal-to-noise ranges of interest. Other approaches, based on the exact calculation of the pairwise error probabilities, are often too cumbersome. A simple exact numerical technique, for calculating, within any desired degree of accuracy, of the pairwise error probability of the HR-STTCM scheme over Rayleigh fading channel is proposed in this dissertation. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2005.

A simple, low cost microwave plasma enhanced chemical vapour deposition (PECVD) technique for the production of carbon nanostructures has been developed in the School‟s Materials Science Laboratory. The technique utilises a conventional microwave oven as an energy source, various hydrocarbons as a carbon source, a metallic aerial as a catalyst and hydrogen to support the process. The input hydrocarbon and the hydrogen flow rate are independently varied to investigate their effect on the resultant nanostructures. This technique allows for the production of carbon nanotubes (CNTs), onion-like nanostructures structures (ONSs) and amorphous carbon, which has been verified via transmission and scanning electron microscopy. A change in input parameters results in the controllable yield of CNTs versus ONSs. The formation of amorphous carbon is reduced by controlling the hydrogen flow rate. In further experiments, the thermal conductivity of the ONSs is investigated using the "Lee‟s Disk" method. It was observed that bulk ONS specimens exhibit a thermal conductivity above that of amorphous carbon powder. Insufficient quantities of CNTs were grown using this method to facilitate a comparable thermal conductivity investigation. / Thesis (M.Sc.)-University of KwaZulu-Natal, Durban, 2010.

The continuous growth in wireless networks is not showing any sign of slowing down as new services, new technologies and new mobile users continue to emerge. Satellite networks are expected to complement the terrestrial network and be a valid option to provide broadband communications services to both fixed and mobile users in scenarios where terrestrial networks cannot be used due to technical and economical viability. In the current emerging satellite networks, where different users with varying traffic demands ranging from multimedia, voice to data and with limited capacity, Radio Resource Management (RRM) is considered as one of the most significant and challenging aspect needed to provide acceptable quality of service that will meet the requirements of the different mobile users. This dissertation considers Packet Scheduling in the Satellite High Speed Downlink Packet Access (S-HSDPA) network. The main focus of this dissertation is to propose a new cross-layer designed packet scheduling scheme, which is one of the functions of RRM, called Queue Aware Channel Based (QACB) Scheduler. The proposed scheduler, which, attempts to sustain the quality of service requirements of different traffic requests, improves the system performance compared to the existing schedulers. The performance analysis comparison of the throughput, delay and fairness is determined through simulations. These metrics have been chosen they are three major performance indices used in wireless communications. Due to long propagation delay in HSDPA via GEO satellite, there is misalignment between the instantaneous channel condition of the mobile user and the one reported to the base station (Node B) in S-HSDPA. This affects effectiveness of the channel based packet schedulers and leads to either under utilization of resource or loss of packets. Hence, this dissertation investigates the effect of the introduction of a Signal-to-Noise (SNR) Margin which is used to mitigate the effect of the long propagation delay on performance of S-HSDPA, and the appropriate SNR margin to be used to achieve the best performance is determined. This is determined using both a semi-analytical and a simulation approach. The results show that the SNR margin of 1.5 dB produces the best performance. Finally, the dissertation investigates the effect of the different Radio Link Control (RLC) Transmission modes which are Acknowledged Mode (AM) and Unacknowledged Mode (UM) as it affects different traffic types and schedulers in S-HSDPA. Proportional fair (PF) scheduler and our proposed, QACB, scheduler have been considered as the schedulers for this investigation. The results show that traffic types are sensitive to the transmitting RLC modes and that the QACB scheduler provides better performance compared to PF scheduler in the two RLC modes considered. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2010.

Electronic noses are targeted at determining odour character in a fashion that emulates conscious odour perception in mammals. The intention of this study was to develop an organisational framework for electronic noses and deploy a sample cheese odour discriminator within this framework. Biological olfactory systems are reviewed with the purpose of extracting the organisational principles that result in successful olfaction. Principles of gas handling, chemoreception, and neural processing are considered in the formulation of an organisational framework. An electronic nose is then developed in accordance with the biologically inspired framework. Gas sensing is implemented by an array of six commercially available (tin oxide) semiconductor sensors. These popular gas sensors are known to lack stability thus necessitating hardware and signal processing measures to limit or compensate for instability. An odorant auto-sampler was developed to deliver measured amounts of odorant to the sensors in a synthetic air medium. Each measurement event encodes a simulated sniff, and is captured across six sensor channels over a period of 256 seconds at a sampling rate of 1Hz. The simulated sniff captures sensor base references and responses to odorant introduction and removal. A technique is presented for representation and processing of sensor-array data as a two-dimensional (2D) image where one dimension encodes time, and the other encodes multi-channel sensory outputs. The near optimal, computationally efficient 2D Discrete Cosine Transform (DCT) is used to represent the 2D signal in a decorrelated frequency domain. Several coefficient selection strategies are proposed and tested. A heuristic technique is developed for the selection of transform domain coefficients as inputs to a non-linear neural network based classifier. The benefits of using the selection heuristic as compared to standard variance-based selection are evident in the results. Benefits include: significant dimensionality reduction with concomitant reduction in classifier size and training time, improved generalisation by the neural network and improved classification performance. The electronic nose produced a 99.1% classification rate across a set of seven different cheeses. / Thesis (M.Sc.Eng.)-University of Natal, Durban, 2003.