Filters
Availability
Publication Year
Resource Type
Faculty
School/Institute
Publication Search Results
Sort
Results per page
1 - 10 of 82
-
(2013)ThesisDue to the general failure of Global Positioning System (GPS) for indoor positioning, non-satellite-based technologies are important for indoor localization. Using wireless networks based on the Received Signal Strength (RSS) location fingerprinting technique is the most popular positioning method used for indoor environments. This research proposes a new positioning technique based on fingerprinting that utilises one of the most available signals of opportunity (SoOP), which is frequency modulated (FM) broadcast radio signals. Then the fusion of FM and Wi-Fi is investigated. The result outperformed the previous methods in terms of accuracy and a more robust and reliable positioning system is presented. Moreover, an analytical framework for estimating the accuracy performance of fingerprinting indoor positioning systems is suggested. Using this model, the most common signal distances such as Euclidean, Manhattan and Chebychev are fully analyses and compared together both mathematically and by simulation so that we can identify which provides least positioning error. Crame-Rao lower bound (CRB) is widely used for assessing localization performance limits but the recent measurement revealed that CRB does not always represent an actual lower bound for indoor positioning. We utilise and modify two more advanced lower bounds and propose an optimization trend in system configuration such that the attained root mean square error in the position estimator gets closer to the minimal attainable variance in the fingerprinting position estimator. Finally, a new method for error estimation in indoor localization systems is designed and novel precision measurements factors for fingerprinting method is developed. Thus the quality of service of the positioning system is improved and the integrity of the system is guaranteed. In this research, the problem of evaluation and enhancement the accuracy of fingerprinting positioning systems utilizing terrestrial FM signals is addressed and analytical frameworks and appropriate solid tools for designing more precise indoor positioning systems are developed. In summary, the FM-based positioning analysis, analytical position error estimation tools, statistical analysis on the accuracy of the indoor positioning systems, and the design criteria tools in this thesis are novel and provide interesting insights into the positioning system performance. These tools are used to optimise the system performance under given performance objectives and constraints.
-
(2010)ThesisAs the backbone for the global communication network, optical dense wavelength division multiplexed (DWDM) systems are facing challenges in capacity, flexibility, reliability and cost effectiveness. In my thesis research I developed five novel optical devices or subsystems to combat these technical challenges. Each of these devices/subsystems is described in an individual chapter including background survey, proposal of new features, theoretical analysis, hardware design, prototype fabrication and characterization, and experimental verification in DWDM systems. The first is a novel tunable asymmetric interleaver that allows the interleaving ratio to be adjusted dynamically. Two design methods were proposed and implemented. Spectral usage optimisation and overall system performance improvement in 10G/40G and 40G/100G systems were successfully demonstrated through simulations and experiments. The second is a colourless intra-channel optical equalizer. It is a passive periodic filter that restores the overall filter passband to a raised cosine profile to suppress the filter narrowing effect and mitigate the inter-symbol interference. 20% passband widening and 40% eye opening were experimentally achieved. The third is a flexible band tunable filter that allows simultaneous tuning of centre frequency and passband width. Based mainly on this filter, a low cost expendable reconfigurable optical add/drop multiplexer (ROADM) node was developed. Its flexible switching features were experimentally demonstrated in a two-ring four-node network testbed. The fourth is a transponder aggregator subsystem for colourless and directionless multi-degree ROADM node. Using the unique characteristics of the coherent receiver, this technology eliminates the requirement of wavelength selector, thus reduces power consumption, size and cost. I experimentally demonstrated that it can achieve < 0.5 dB penalty between receiving single channel and 96 channels. The last is a real-time feedforward all-order polarization mode dispersion (PMD) compensator. It first analyses spectral interference pattern to retrieve phase information and calculate PMD, then it uses a pulse shaper to restore the pulse shape and thus compensates the PMD. These functions were demonstrated through experiments and simulations. All of these novel devices and subsystems deliver new functional features and are suitable to be applied in the next generation DWDM systems to improve capacity, flexibility, and reliability and to reduce cost.
-
(2011)ThesisThis dissertation presents a novel prototype optical interconnect system achieved by MEMS technology for stacked silicon dies. The overall approach is drawn from concepts on 3D-TSV metal interconnect and introducing an optical interconnect. Optical interconnect is a promising solution for future global interconnect inside a chip package consisting of a stack of multiple dies. The design allows an optical beam to propagate both in plane and out-of-plane, hence establishing optical communication between stacked dies. The entire system includes a pair of 45 micromirrors with one in an upward-facing (front) position and the other in downward-facing (rear) position. Fiber grooves were fabricated on each dies for the purpose of optical testing. Passive alignment structures have been integrated in the system to achieve accurate horizontal alignment. The research work have also proposed a planar silica graded index lens integrated 3D optical interconnect system. A pair of planar silica graded index lens is designed to sit in front of the 45 micromirror pair to mitigate the free space propagation loss. Optical beam free space spot sizes from these planar lenses are commensurate with the micromirror dimensions. To fabricate high quality ultra-smooth 45 micromirror on the (100) monocrystalline silicon surface by anisotropic wet etching, surfactant added TMAH etching behavior was first investigated. Non-ionic surfactants NCW-601A, Triton X-100, NCW-1002 with both high concentration (25%) and low concentration (10%) TMAH solutions was extensively studied for their etching properties on (100) and (110) silicon surfaces. Etch rate, selectivity and surface roughness were studied. Etchant temperature was also varied from 60C to 90C to see the effects on etching behaviour. This research is the first reported use of a new surfactant NCW-1002 in low concentration TMAH. Better selectivity and surface roughness of the (110) plane (i.e. micromirror surface) can be achieved with this surfactant added low concentration TMAH at 60C. In view of the intrinsic partial isotropic etching behavior in the vicinity of the (100) plane for low concentration TMAH with high concentration surfactant, this work has established and demonstrated for the first time in published literature that only a limited portion of the micromirror is truly 45 and is located approximately at the middle of the slope. To reduce the top curved portion, two new techniques have been developed. One technique involves timed HF dip treatments during the etching process to remove the overhanging oxide, which has impeded the etching at the top part of the slope due to the diffusion-dependant effect. The other technique involves a maskless etching. The entire top curved slope can thus be straightened out. To avoid the bottom curved portion, another technique was developed to allow a deeper etch at the lower end of the micromirror to recess the non-45 lower portion of the slope. Both these techniques can nearly double the size of the truly 45 portion of the micromirror. Based on the framework of a front upward-facing 45 micromirror, a processing technique was developed to fabricate a novel rear downward-facing 45 micromirror using electro-chemical TMAH release. With the assistance of passive alignment structures, optical test was conducted to confirm successful optical interconnect between stacked dies using optical fiber. The loss was measured to be 14.7dB, among which free space transmission loss is approximately 10dB. To reduce the free space propagation loss, a pair of planar silica graded index lens integrated optical interconnect system was then proposed. Fabrication of the planar silica graded index lens with front 45 micromirror is confirmed in this research work.
-
(2013)ThesisBeamforming in wireless communications have gathered great interests in recent years due to its ability to enhanced the performance of networks significantly by exploiting intensively the spatial diversity. In this work, the objectives of beamforming design consist of several optimization targets ranging from minimizing the beamforming power subject to quality-of-service (QoS) constraints to maximizing the minimum QoS regarding fixed budgets of transmitting power. The design problems of beamforming are intrinsically challenging because their natural formulations are nonconvex optimization problems. Moreover several problems are proved to be non-deterministic polynomial-time hard (NP-hard) such as beamforming in multicast transmission. These problems are very difficult to solve at optimality in practical sense. Therefore, there is a strong motivation to convert the original design problems into a series of convex problems with desirable computational complexity by applying efficient optimization techniques. This dissertation contributes mainly in exploiting the convex optimization algorithms to solve nonconvex beamforming problems in several network settings. First, the transmit beamforming for downlink communication of multicast transmission with spectrum sharing is investigated. Secondly, the beamforming design is applied on amplify-forward (AF) wireless relaying systems using single-antenna relays. The key contribution is to derive the beamforming schemes applied on transmit antennas so that the beamforming power is minimized while all users' signal-to-interference-and-noise ratios (SINRs) are guaranteed. The formulation results in nonconvex optimization problems due to SINR constraints hence require to be converted into semidefinite programming (SDP) forms. The SDP problems are again nonconvex regarding the rank-one constraints on semidefinite variables. Conventionally, the rank-one constraints are relaxed hence the problems cannot be solved thoroughly. In this work, nonsmooth optimization techniques are employed to tackle with the nonconvex rank-one constraints and are successfully to deliver efficient solutions that can outperform the conventional methods. Finally the precoding design problems in mutiple-input multiple-output (MIMO) relaying scenarios are considered. The difference-of-two-convex-function (D.C.) programming technique is employed to solve the problems at optimality with significantly lower complexity compared with conventional method.
-
(2012)ThesisFuel cells (FCs) are a promising alternative energy solution and a reliable technology for niche applications providing high conversion efficiency. FCs are generally classified according to the type of the electrolyte. Proton exchange membrane FCs (PEMFCs) combine the hydrogen fuel with the oxygen from the atmosphere to produce water, heat (at temperature up to 90°C) and electricity. The major advantages of the PEMFCs include: being environmentally friendly, offering high efficiency and low operating temperature while providing increased safety. When low-voltage unregulated FC output is conditioned to generate AC power, two stages are generally required: a boost stage and an inversion one. However, due to the cascaded power stage conventional two-stage FC systems have unavoidable drawbacks such as being bulky, inefficient and having a high cost. Moreover, FC systems require an auxiliary unit to improve efficiency, lifetime and safety by cancelling the ripple current and supporting its slow dynamics. In this thesis, single-stage FC systems are reported to overcome the disadvantages of the conventional FC systems caused by cascaded power stages. The proposed FC systems provide high power conversion efficiency, low-cost and compactness for standalone and grid connected applications in single- and three-phase systems. Analysis, simulation and experimental results taken from 1kW laboratory prototypes operating at 20 kHz are presented to validate the performance of the proposed FC systems. Single conversion stage FC systems based on the boost-inverter and buck-boost inverter with a back-up energy storage unit have been developed in this research as follows: • A single power stage FC system based on a boost-inverter with a battery back-up unit • A grid-connected FC system based on a boost-inverter with a battery back-up unit • A single conversion stage FC system based on a buck-boost inverter with a battery back-up unit • A single conversion stage three-phase FC system based on a boost-inverter with a battery back-up unit • Design and implementation of digital control as implemented in an FC energy system The proposed FC systems has achieved satisfactory performance for delivering boosting and inversion functions within the single power conversion offering high efficiency and compactness. Moreover, an integrated battery back-up energy storage unit has also performed two functions: providing the ramping operation to deal with the slow dynamics of the FC and eliminating the ripple current to increase the efficiency and life time of the FC.
-
(2012)ThesisGrowing concerns over high and volatile fossil-fuel prices, energy security and climate change present significant sustainability challenges for electricity industries around the world. They have particular implications for generation investment and planning as decisions to build power plants represent major commitments with regard to future electricity industry costs, fuel dependencies, energy security outcomes, and greenhouse emissions. Increasing uncertainty about future fuel prices, plant construction costs, climate change policies and electricity demand adds to the challenges. In consequence, generation investment decision-making is transitioning from traditional cost minimisation approaches towards more complex assessments incorporating future uncertainties and a range of industry objectives. This thesis presents a novel generation investment and planning decision-support tool for assessing possible future generation portfolios within an electricity industry under uncertainty and multiple policy objectives. The tool extends conventional optimal generation mix concepts by using Monte Carlo simulation and portfolio analysis techniques to determine probability distributions of future expected generation costs and greenhouse emissions for a wide range of generation portfolios. The tool can incorporate complex and correlated probability distributions for future fossil-fuel prices, carbon prices, plant investment costs, and electricity demand including price elasticity impacts. It supports sophisticated risk assessments, including downside economic risks, for any generation portfolio. A post-processing analysis is also implemented to assess the potential operational dispatch constraints and costs associated with different portfolios in meeting time-varying demand. Applications of this tool are demonstrated through case studies of electricity industries with different generation options facing a range of future uncertainties. The results provide high-level insights into issues that will likely be key drivers of future industry performance such as the impact of different sources of uncertainty, the role of future demand changes in response to electricity prices, and the interaction between wind penetration and carbon pricing on the expected costs, cost risks, and emissions of different portfolios. The implications of nuclear power and the value of different technologies within generation portfolios are also explored. These results highlight the potential value of the tool in facilitating utility and policy decision-making in the challenging task of generation investment and planning for increasingly uncertain future electricity industries.
-
(2012)ThesisIn recent years, optical fibre laser-based sensors have attracted great research attention. Numerous types of optical fibre laser-based sensors have been demonstrated for various applications with desirable features such as high signal power, low noise, small size, inherited fibre compatibility, single polarization and single longitudinal mode operation. In particular, fibre laser-based sensors with high spectral resolution are appropriate for high sensitivity and low concentration gas detection applications. This thesis reports the research work on a fibre laser-based intra-cavity absorption gas sensor (ICAGS). The ICAGS has many advantages such as long interaction length, compact configuration and capability of multiplexing and distributed measurement. A few key technical issues related to modeling, design, evaluation and fabrication of a fibre-ring laser based ICAGS were systematically investigated. The system technologies, parameters, conditions, considerations and their impacts on the ICAGS based on an erbium doped fibre (EDF) were investigated theoretically and experimentally. Firstly we studied in great detail the effects of key system parameters on the performance of the ICAGS. Based on the rate equations and propagation equations, a theoretical model to evaluate the output characteristics of the ICAGS was established. For the first time, the relations between the key design parameters (EDF dopant concentration, EDF length, cavity loss and pump power) and the system performance parameters, e.g. sensitivity enhancement factor, of an ICAGS were systematically studied. Especially, appropriate selections of the design parameters, i.e. EDF factor (the product of EDF dopant concentration and EDF length), cavity loss and pump power, for optimal performance in terms of sensitivity enhancement and signal output power were obtained. Secondly we established an EDF based ICAGS experimental system. Detailed system considerations, hardware selections and software designs are presented. The nonlinearity of the wavelength selective element - Fabry-Perot tunable filter, was investigated. A method for compensating the effects of the nonlinearity was proposed. This method reduces the error in identifying the central wavelength of the absorption lines and increases the accuracy of the central wavelength detection effectively without a feedback system. Thirdly we proposed and analyzed a new design of miniature gas cell to solve the problems of previous gas cells in extreme environment applications. This gas cell was made of a pair of gradient-index fibre lenses and a narrow capillary which made it compact and stable. Good coupling efficiency could be achieved if proper gradient-index fibre is chosen. Theoretical analysis and experiments were carried out to prove the feasibility of the design. Fourthly experiments on gas sensing using a wavelength sweep technique were conducted. By selecting appropriate system operational parameters, 6 strong absorption lines near 1530 nm of acetylene can be obtained with good spectrum resolution in one scanning period. For various concentration cases, by choosing proper absorption lines and performing weighted averaging, detection accuracy can be enhanced. To further improve the accuracy in determining the central wavelength of the absorption lines and to realize multi-gases recognition, a fibre Bragg grating (FBG) as the wavelength reference has been introduced. Finally, potential applications of micro-structured fibre (MOF) in an ICAGS were explored. Optical fibre gas sensors using MOF open up new opportunities for exploiting the interaction of light with gases and exhibit great sensing potential. The splicing techniques of MOF to SMF were studied using a commercial fusion splicer. With proper splicing parameters, a splice loss of about 1.62 dB was realized by a commercial splicer.
-
(2013)ThesisReal time communications have, in the last decade, become a highly relevant component of Internet applications and services, with both interactive communications and streamed content being used in developed and developing countries alike. Due to the proliferation of mobile devices, wireless media is becoming the means of transmitting a large part of this increasingly important real time communications traffic. Wireless has also become an important technology in developing countries, with satellite communications being increasingly deployed for traffic backhaul and ubiquitous connection to the Internet. A number of issues need to be addressed in order to have an acceptable service quality for real time communications in wireless environments. In addition to this, the availability of multiple wireless interfaces on mobile devices presents an opportunity to improve and further exacerbates the issues already present on single wireless links. Therefore in this thesis, we consider improvements to transport protocols for real time communications and streaming services to address these problems and we provide the following contributions. To deal with wireless link issues of errors and delay, we propose two enhancements. First, an improvement technique for Datagram Congestion Control Protocol CCID4 for long delay wireless (e.g. satellite) links, demonstrating significant performance improvements for Voice over IP applications. To deal with link errors, we have proposed, implemented and evaluated an erasure coding based packet error correction approach for Concurrent Multipath Transfer extension of Stream Control Transport Protocol data transport over multiple wireless paths. We have identified packet reordering as a major cause of performance degradation in both single and multi-path transport protocols for real time communications and media streaming. We have proposed a dynamically resizable buffer based solution to mitigate this problem within the DCCP protocol. For improving the performance of multi-path transport protocols over dissimilar network paths, we have proposed a delay aware packet scheduling scheme, which significantly improves the performance of multimedia and bulk data transfer with CMT-SCTP in heterogeneous multi-path network scenarios. Finally, we have developed a tool for online streaming video quality evaluation experiments, comprising a real-time cross-layer video streaming technique implemented within an open-source H.264 video encoder tool called x264.
-
(2011)ThesisAccelerometry shows promise in providing an inexpensive but effective means of understanding the human gait. Accurate classification of everyday gait patterns could allow a monitoring system to exhibit greater ‘intelligence’, like, improving the ability to detect and monitor the activities of a person, in a more accurate tracking of the person health parameters in an unsupervised environment. This thesis develops a gait classification algorithm using features derived from novel gait models. The Linear Predictive (LP) model is proposed to model the gait with the basic assumption that the gait is a system that comprises impulse trains input, which correspond to the footstrike. From the Linear Predictive model, the Linear Predictive Cepstral Coefficients (LPCC) were proposed as features, as they have better class separation compared to using LP coefficients directly. The filterbank energies were proposed to approximate the LPCC with a smaller number of parameters. Spectral Centroid Amplitude and Spectral Centroid Frequency features were also developed to integrate frequency information with the proposed filterbank features. The second gait model that is proposed in this thesis is a harmonic model. The premise behind this model is that the accelerometer signal’s spectral contains a fundamental frequency found to be the walking stride rate with multiple harmonics that fit a harmonic model. The first four harmonics were found to be a good approximation of the acceleration movement of the hip. The magnitudes of the harmonics were used for classification features and proved to be better features than the linear predictive filterbank features. Another advantage of the harmonic features is that they are independent of walking speed, which therefore enables the extraction of features that are robust to the variations of speed. Novel linear delta zero crossing counts regression features were used as complimentary features to the proposed static features from the model parameters. The assumption that the zero crossing counts provide a good characterisation of the amount of artefacts caused by the muscle movements was the motivation for using these features. Finally, the non-linear Empirical Mode Decomposition (EMD) method was investigated to extract nonlinear features for gait pattern classification. The Bayesian adapted GMM classification system was used as a back-end system to further improve the classification accuracy of the system by removing the subject variation of the system. In addition, score level fusion was proposed to integrate the linear regression delta zero crossing features into the static features.
-
(2010)ThesisThis thesis is concerned with the problem of robust model predictive control (MPC) of an input and state constrained linear system, for the case that descriptions of uncertainty, present within the control loop, are time-variant. Such descriptions shall be referred to as time-varying uncertainty descriptions in this thesis. The study of such problems is motivated by control applications, for which the sets, within which the uncertainty is known to evolve in, are time-variant, and for which future predictions of such uncertainty description sets from previous times might need to be revised at subsequent times, owing to additional information becoming available in the future. Within the framework of robust MPC subject to time-varying uncertainty descriptions, the contribution of this thesis is twofold, as outlined next. Firstly, a robust MPC controller is proposed, which ensures robustness, performance, constraint satisfaction and stability, all in a suitably defined sense, under time-varying uncertainty descriptions, which are assumed to satisfy a mild condition of consistency. The key idea behind the proposed controller consists in the regulation of the system state around an "optimized" average state, such that the system state is ensured to evolve "tightly" around the optimized average state. The obtained results are shown to represent an extension of works from two related streams of the robust MPC literature (the tube based and constraint restriction based robust MPC streams). It is shown through simulations that the proposed controller, even for the case of time-invariant uncertainty descriptions, outperforms previous robust MPC methods, owing to a decreased degree of conservatism, in a suitably defined sense, over previous methods, but also owing to its ability to ensure improved uncertainty rejection, over previous methods, through " tighter" regulation of the actual state around the optimized average state through employment of a "feedback corrective term", as part of the proposed control law. The second contribution of this thesis relates to the analysis of the computational complexity of the proposed control algorithm. It is shown that problems with the computational complexity encountered in previous robust IVIPC methods, associated with computations of Minkowski sums and Minkowski differences of sets, may be overcome, by employing the theory of support functions and support vectors, which have been studied extensively in the related field of control theory dealing with constructions of reachable sets. In particular, it is shown that the computational complexity may either be broken down into simpler subproblems, or that further simplifications in computational complexity may be achieved, if additional structure, as motivated by practically relevant control problems, is assumed on the time-varying uncertainty descriptions. The developed theory of this thesis is compared against previous methods of relevance, where it is shown that, owing to the particular formulation of the control law proposed in this thesis, improved performance is achieved, even for the case that time-invariant uncertainty descriptions are assumed. The theory is furthermore applied to two practical problems, both of which serve to illustrate the validity and utility of the developed theory. Firstly, the theory is applied to the problem of robust MPC, when an observer must be employed in order to estimate the system state. Subsequently, the theory is applied to the problem of control over a communication channel of limited capacity and with data losses.