published_or_final_version / abstract / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / abstract / Mechanical Engineering / Master / Master of Philosophy

Our thesis work focuses on Frequency-domain Blind Source Separation (BSS) in which the received mixed signals are converted into the frequency domain and Independent Component Analysis (ICA) is applied to instantaneous mixtures at each frequency bin. Computational complexity is also reduced by using this method. We also investigate the famous problem associated with Frequency-Domain Blind Source Separation using ICA referred to as the Permutation and Scaling ambiguities, using methods proposed by some researchers. This is our main target in this project; to solve the permutation and scaling ambiguities in real time applications / Gozie: modebelu2001@yahoo.com Anayo: ezeudea@yahoo.com

The gap between abstract, mathematics-oriented research in cryptography and the engineering approach of designing practical, network security protocols is widening. Network researchers experiment with well-known cryptographic protocols suitable for different network models. On the other hand, researchers inclined toward theory often design cryptographic schemes without considering the practical network constraints. The goal of this dissertation is to address problems in these two challenging areas: building bridges between practical network security protocols and theoretical cryptography. This dissertation presents techniques for building performance sensitive security protocols, using primitives from linear feedback register sequences (LFSR) sequences, for a variety of challenging networking applications. The significant contributions of this thesis are: 1. A common problem faced by large-scale multicast applications, like real-time news feeds, is collecting authenticated feedback from the intended recipients. We design an efficient, scalable, and fault-tolerant technique for combining multiple signed acknowledgments into a single compact one and observe that most signatures (based on the discrete logarithm problem) used in previous protocols do not result in a scalable solution to the problem. 2. We propose a technique to authenticate on-demand source routing protocols in resource-constrained wireless mobile ad-hoc networks. We develop a single-round multisignature that requires no prior cooperation among nodes to construct the multisignature and supports authentication of cached routes. 3. We propose an efficient and scalable aggregate signature, tailored for applications like building efficient certificate chains, authenticating distributed and adaptive content management systems and securing path-vector routing protocols. 4. We observe that blind signatures could form critical building blocks of privacypreserving accountability systems, where an authority needs to vouch for the legitimacy of a message but the ownership of the message should be kept secret from the authority. We propose an efficient blind signature that can serve as a protocol building block for performance sensitive, accountability systems. All special forms digital signatures—aggregate, multi-, and blind signatures—proposed in this dissertation are the first to be constructed using LFSR sequences. Our detailed cost analysis shows that for a desired level of security, the proposed signatures outperformed existing protocols in computation cost, number of communication rounds and storage overhead.

Consumers evaluate products in the market place using their senses and often form mental representations of product properties. These mental representations have been studied extensively. Imagery has been shown to interact with perception within many perceptual modalities including vision, auditory, olfactory, and motor. This dissertation draws on the vast visual imagery literature to examine imagery in the haptic, or touch, modality. Two studies were undertaken to examine the relationship between haptic imagery and haptic perception The first study is based on studies from cognitive psychology that have used similar methods for examining visual imagery and visual perception. In study 1, sighted and visually impaired participants were asked to evaluate objects haptically, to form a haptic image of that object during a short interval, and then to compare the haptic image to a second object. In Study 2, sighted and visually impaired participants listened to five radio advertisements containing imagery phrases from multiple modalities. After listening to the advertisements, participants were asked to recall the ad content and assess both the ad and the product while haptically evaluating the product in the ad. Though results were mixed and further exploration will be necessary, these studies offer broad implications for consumer use of haptic imagery in shopping environments. The implications for both sighted and blind consumers are discussed.

High quality digital images have become pervasive in modern scientific and everyday life — in areas from photography to astronomy, CCTV, microscopy, and medical imaging. However there are always limits to the quality of these images due to uncertainty and imprecision in the measurement systems. Modern signal processing methods offer the promise of overcoming some of these problems by postprocessing these blurred and noisy images. In this thesis, novel methods using nonstationary statistical models are developed for the removal of blurs from out of focus and other types of degraded photographic images. The work tackles the fundamental problem blind image deconvolution (BID); its goal is to restore a sharp image from a blurred observation when the blur itself is completely unknown. This is a “doubly illposed” problem — extreme lack of information must be countered by strong prior constraints about sensible types of solution. In this work, the hierarchical Bayesian methodology is used as a robust and versatile framework to impart the required prior knowledge. The thesis is arranged in two parts. In the first part, the BID problem is reviewed, along with techniques and models for its solution. Observation models are developed, with an emphasis on photographic restoration, concluding with a discussion of how these are reduced to the common linear spatially-invariant (LSI) convolutional model. Classical methods for the solution of illposed problems are summarised to provide a foundation for the main theoretical ideas that will be used under the Bayesian framework. This is followed by an indepth review and discussion of the various prior image and blur models appearing in the literature, and then their applications to solving the problem with both Bayesian and nonBayesian techniques. The second part covers novel restoration methods, making use of the theory presented in Part I. Firstly, two new nonstationary image models are presented. The first models local variance in the image, and the second extends this with locally adaptive noncausal autoregressive (AR) texture estimation and local mean components. These models allow for recovery of image details including edges and texture, whilst preserving smooth regions. Most existing methods do not model the boundary conditions correctly for deblurring of natural photographs, and a Chapter is devoted to exploring Bayesian solutions to this topic. Due to the complexity of the models used and the problem itself, there are many challenges which must be overcome for tractable inference. Using the new models, three different inference strategies are investigated: firstly using the Bayesian maximum marginalised a posteriori (MMAP) method with deterministic optimisation; proceeding with the stochastic methods of variational Bayesian (VB) distribution approximation, and simulation of the posterior distribution using the Gibbs sampler. Of these, we find the Gibbs sampler to be the most effective way to deal with a variety of different types of unknown blurs. Along the way, details are given of the numerical strategies developed to give accurate results and to accelerate performance. Finally, the thesis demonstrates state of the art results in blind restoration of synthetic and real degraded images, such as recovering details in out of focus photographs.

Singing voice separation (SVS) can be defined as the process of extracting the vocal element from a given song recording. The impetus for research in this area is mainly that of facilitating certain important applications of music information retrieval (MIR) such as lyrics recognition, singer identification, and melody extraction. To date, the research in the field of SVS has been relatively limited, and mainly focused on the extraction of vocals from monophonic sources. The general approach in this scenario has been one of considering SVS as a blind source separation (BSS) problem. Given the inherent diversity of music, such an approach is motivated by the quest for a generic solution. However, it does not allow the exploitation of prior information, regarding the way in which commercial music is produced. To this end, investigations are conducted into effective methods for unsupervised separation of singing voice from stereophonic studio recordings. The work involves extensive literature review of existing methods that relate to SVS, as well as commercial approaches. Following the identification of shortcomings of the conventional methods, two novel approaches are developed for the purpose of SVS. These approaches, termed SEMANICS and SEMANTICS draw their motivation from statistical as well as spectral properties of the target signal and focus on the separation of voice in the frequency domain. In addition, a third method, named Hybrid SEMANTICS, is introduced that addresses time‐, as well as frequency‐domain separation. As there is lack of a concrete standardised music database that includes a large number of songs, a dataset is created using conventional stereophonic mixing methods. Using this database, and based on widely adopted objective metrics, the effectiveness of the proposed methods has been evaluated through thorough experimental investigations.

Qatar is a small country in the Middle East which has used its oil wealth to invest in the country's infrastructure and education. The technology for Internet voting now exists or can be developed, but are the people of Qatar willing to take part in Internet voting for national elections?. This research identifies the willingness of government and citizens to introduce and participate in Internet voting (I-voting) in Qatar and the barriers that may be encountered when doing so. A secure I voting model for the Qatar government is then proposed that address issues of I-voting which might arise due to the introduction of such new technology. Recommendations are made for the Qatar government to assist in the introduction of I-voting. The research identifies the feasibility of I-voting and the government s readiness and willingness to introduce it. Multiple factors are examined: the voting experience, educational development, telecommunication development, the large number of Internet users, Qatar law which does not bar the use of I-voting and Qatar culture which supports I-voting introduction. It is shown that there is a willingness amongst both the people and the government to introduce I-voting, and there is appropriate accessibility, availability of IT infrastructure, availability of Internet law to protect online consumers and the existence of the e government project. However, many Qataris have concerns of security, privacy, usability, transparency and other issues that would need to be addressed before any voting system could be considered to be a quality system in the eyes of the voters. Also, the need to consider the security threat associated on client-side machines is identified where a lack of user awareness on information security is an important factor. The proposed model attempts to satisfy voting principles, introducing a secure platform for I-voting using best practices and solutions such as the smart card, Public Key Infrastructure (PKI) and digital certificates. The model was reviewed by a number of experts on Information Technology, and the Qatari culture and law who found that the system would, generally, satisfy voting principles, but pointed out the need to consider the scalability of the model, the possible cyber-attacks and the risks associated with voters computers. which could be reduced by enhancing user awareness on security and using secure operating systems or Internet browsers. From these findings, a set of recommendations were proposed to encourage the government to introduce I-voting which consider different aspects of I-voting, including the digital divide, e-literacy, I voting infrastructure, legal aspects, transparency, security and privacy. These recommendations were also reviewed by experts who found them to be both valuable and effective. Since literature on Internet voting in Qatar is sparse, empirical and non-empirical studies were carried out in a variety of surveys, interviews and experiments. The research successfully achieved its aim and objectives and is now being considered by the Qatari Government.

This research project was a descriptive analysis of the computer usage and skills of academic students in grades 10-12 who are legally blind and attending public school in the Region 10 Education Service Center service area of Texas. In addition, this study provided a process that other regions in the state or educational agencies may duplicate to document the computer skills and usage of students with visual impairments in their area. Twenty-seven students who are legally blind were surveyed by their teachers of the visually impaired regarding their computer usage and skill abilities, and eleven of the twenty-seven students were interviewed by the researcher to gain further information pertaining to computer usage and future plans upon graduation. Using prior research as a basis for understanding how sighted students used the computer, it was found that students who are legally blind used the computer similarly to their sighted peers except that students with significant visual impairments seemed to use to the computer to listen to music more than their sighted counterparts. In addition, students who are legally blind indicated that they learned most of their computer skills at school rather than at home like their sighted teenagers. Furthermore, it was determined that students who are legally blind were not learning the computer skills necessary for success in post-secondary education and vocational endeavors. Although the students were being exposed to many different computer applications, most did not use the applications weekly, nor report that they were experienced with the majority of basic skills related to applications such as word processing, Internet searching, emailing, spreadsheets and databases.

An increasing amount of information content used in schools, work and everyday living is being presented in graphical form, creating accessibility challenges for individuals who are blind or visually impaired, especially in dynamic environments, such as over the internet. Refreshable haptic displays that can interact with computers can be used to access such information tactually. Main focus of this study was the development of specialized computer applications allowing users to actively compensate for the inherent issues of haptics when exploring visual diagrams as compared to vision, which we hypothesized, would improve the usability of such devices. An intuitive zooming algorithm capable of automatically detecting significant different zoom levels, providing auditory feedback, preventing cropping of information and preventing zooming in on areas where no features were present was developed to compensate for the lower spatial resolution of haptics and was found to significantly improve the performance of the participants. Another application allowing the users to perform dynamic simplifications on the diagram to compensate for the serial based nature of processing 2D geometric information was tested and found to significantly improve the performance of the participants. For both applications participants liked the user interface and found it more usable, as expected. In addition, in this study we investigated methods that can be used to effectively present different visual features as well as overlaying features present in the visual diagrams. Three methods using several combinations of tactile and auditory modalities were tested. We found that the performance significantly improves when using the overlapping method using different modalities. For tactile only methods developed for deaf blind individuals, the toggle method was surprisingly preferred as compared to the overlapping method.