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

published_or_final_version / abstract / Electrical and Electronic Engineering / Master / Master of Philosophy

published_or_final_version / abstract / Industrial and Manufacturing Systems Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / Electrical and Electronic Engineering / Master / Master of Philosophy

The main thrust of this dissertation is the application of statistics and information theory to design, analysis and estimation pertaining to image-forming systems. This study explores the application of Shannon's information in pupil design, the characterization of noise, and study of its behavior in a specific electro-optical system, and estimation of the degraded spread function in atmospherical imagery using the maximum entropy method. Our study shows that a pupil designed to maximize Shannon's information throughput is an apodizer, resulting in resolution and contrast enhancement when compared to the diffraction-limited case. The Strehl ratio is about 0.55. Investigation of statistical and spectral properties as a function of gray level in an electro-optical tracking system indicates that the noise is "white," having a wide band and a close-to-Gaussian distribution. Estimating the spread function via maximum entropy technique has revealed some remarkable results. Using an edge as the object, simulation studies predict a superior estimate in the mean squared error sense to those of the least squares in the presence of three types of noise (signal-dependent Gaussian and Poisson, and signal-independent Gaussian noise). Information theory, linear systems theory, sampling theory and more particularly, statistics and the Fast Fourier Transform are used to derive our results.

This paper describes the design and simulation of a network interface unit (NIU) for a picture archiving and communication system (PACS) network called PACnet. PACnet is a dual fiber optic ring network under development at the Computer Engineering Research Laboratory of the University of Arizona. This network integrates voice, data, and image communications in a hospital environment and supports a throughput rate between 200-500 megabits per second. At each node in the network, an NIU implements the Data Link Layer and Physical Layer protocols of PACnet. The initial network interface unit design for PACnet was a functional description of NIU protocols and major components. In order to construct a demonstration prototype of PACnet,the NIU description must be refined and an architecture must be specified. The NIU design is specified and simulated using the hardware description language VHDL. Simulation results provide information on NIU timing characteristics and logic families required to implement the NIU.

<p>Web technologies have necessitated a transformation culture in higher education<br /> institutions. Many of these institutions are employing web technologies whose<br /> development, for varying reasons, is not supported by research in their field and<br /> domain of use. One such field is instructional design for the web learning.<br /> Although there is a lot of research on the most effective instructional design<br /> strategies, the use of research for web-based learning applications has been<br /> limited. <br /> <br /> This thesis reports on a study aimed at transforming the research on instructional<br /> design into practice by designing an instructional design system and providing an<br /> argument for its implementation. The argument is intended to facilitate the design<br /> and development of an instructional design subsystem of the web, that would in<br /> turn offer effective and efficient ways for creating web-based learning materials<br /> to instructors.<br /> <br /> The study started by examining the various paradigms, theories and practices of<br /> instructional design with the intent of using them to enrich and improve the<br /> practice of instructional design in web learning. It undertook a thorough and<br /> systematic review of the literature on instructional design in order to come up<br /> with an instructional design system. The design approach used successful design<br /> patterns that have been used elsewhere, e.g. in software design, to create common<br /> responses or solutions to recurrent problems and circumstances. Instructional<br /> design patterns were identified in this study as the recurrent problems or processes<br /> instructional designers go through while creating instructional materials, whose<br /> solutions can be reused over and over again.<br /> <br /> This study used an iterative developmental research process of finding and<br /> modelling an instructional design process as the research methodology. This<br /> process follows and builds on existing research on instructional models, theories<br /> and strategies, and ensures that the same methodology can be used to test the<br /> theories in the design, thus improving both the research and the design.</p>

The main crux of this thesis was to produce a model that was capable of simulating the theoretical performance of different configurations for a spatial measurement system using radio frequency technology. It has been important to study new modalities of spatial measurement since spatial measurement systems are an enabling technology that have allowed for the creation of better medical procedures and techniques, provided valuable data for motion capture in animation and biomechanics, and have improved the quality of manufacturing processes in many industries. However, there has been room for improvement in the functional design and accuracy of spatial measurement systems that will enhance current applications and further develop new applications in medicine, research and industry. <br /><br /> In this thesis, a modelling framework for the investigation of spatial measurement based on radio frequency signals was developed. The simulation framework was designed for the purpose of investigating different position determination algorithms and sensor geomatries. A finite element model using the FEMLAB partial differential equation modelling tool was created for a time-domain model of electromagnetic wave propagation in order to simulate the radio frequency signals travelling from a transmitting source antenna to a set of receiving antenna sensors. Electronic line signals were obtained using a simple receiving infinitesimal dipole model and input into a time difference of arrival localization algorithm. The finite element model results were validated against a set of analytical solutions for the free space case. The accuracy of the localization algorithm was measured against a set of possible applications for a potential radio frequency spatial measurement system design. <br /><br /> It was concluded that the simulation framework was successful should one significant deficiency be corrected in future research endeavours. A phase error was observed in the signals extracted at the receiving antenna locations. This phase error, which can be up to 40??, was attributed to the zeroth order finite elements implemented in the finite element model. This phase error can be corrected in the future if higher order vector elements are introduced into future versions of FEMLAB or via the development of custom finite element analysis software but were not implemented in this thesis due to time constraints. Other improvements were also suggested for future work.

Theoretical procedures are developed to account for the effect of emotion and perception in strategic conflict. The <em>possibility principle</em> facilitates modeling the effects of emotions on future scenarios contemplated by decision makers; <em>perceptual graph models</em> and a <em>graph model system</em> permit the decision makers (DMs) to experience and view the conflict independently; and <em>perceptual stability analysis</em>, which is based on individual- and meta-stability analysis techniques, is employed in analyzing graph model systems when the DMs have inconsistent perceptions. These developments improve the methodology of the Graph Model for Conflict Resolution by reconciling emotion, perception, and strategy to make predictions consistent with the actual unfolding of events. <br /><br /> Current research in neuroscience suggests that emotions are a necessary component of cognitive processes such as memory, attention, and reasoning. The somatic marker hypothesis, for example, holds that feelings are necessary to reasoning, especially during social interactions (Damasio, 1994, 2003). Somatic markers are memories of past emotions: we use them to predict future outcomes. To incorporate the effect of emotion in conflict, the underlying principle of Damasio?s hypothesis is used in developing the possibility principle, which significantly expands the paradigm of the Graph Model for Conflict Resolution of Fang, Hipel, and Kilgour (1993). <br /><br /> State identification is a crucial step in determining future scenarios for DMs. The possibility principle is integrated into the modeling stage of the Graph Model by refining the method of determining feasible states. The possibility principle enables analysts and DMs to include emotion in a conflict model, without sacrificing the parsimonious design of the Graph Model methodology, by focusing attention on two subsets of the set of feasible states: <em>hidden</em> and <em>potential</em> states. Hidden states are logically valid, feasible states that are invisible because of the presence of negative emotions such as anger and fear; potential states are logically valid, feasible states that are invisible because of missing positive emotions. Dissipating negative emotions will make the hidden states visible, while expressing the appropriate positive emotions will make the potential states visible. The possibility principle has been applied to a number of real world conflicts. In all cases, eliminating logically valid states not envisioned by any DM simplifies a conflict model substantially, expedites the analysis, and makes it an intuitive and a realistic description of the DMs' conceptualizations of the conflict. <br /><br /> A fundamental principle of the Graph Model methodology is that all DMs' directed graphs must have the same set of feasible states, which are integrated into a <em>standard</em> graph model. The possibility principle may modify the set of feasible states perceived by each DM according to his or her emotion, making it impossible to construct a single standard graph model. When logically valid states are no longer achievable for one or more DMs due to emotions, the apprehension of conflict becomes inconsistent, and resolution may become difficult to predict. Therefore, reconciling emotion and strategy requires that different apprehensions of the underlying decision problem be permitted, which can be accomplished using a perceptual graph model for each DM. A perceptual graph model inherits its primitive ingredients from a standard graph model, but reflects a DM's emotion and perception with no assumption of complete knowledge of other DMs' perceptions. <br /><br /> Each DM's perceptual graph model constitutes a complete standard graph model. Hence, conclusions drawn from a perceptual graph model provide a limited view of equilibria and predicted resolutions. A graph model system, which consists of a list of DMs' perceptual graph models, is defined to reconcile perceptions while facilitating conclusions that reflect each DM's viewpoint. However, since a DM may or may not be aware that other graph models differ from his or her own, different variants of graph model systems are required to describe conflicts. Each variant of graph model system corresponds to a configuration of awareness, which is a set of ordered combinations of DMs' viewpoints. <br /><br /> Perceptual stability analysis is a new procedure that applies to graph model systems. Its objective is to help an <em>outside</em> analyst predict possible resolutions, and gauge the robustness and sustainability of these predictions. Perceptual stability analysis takes a two-phase approach. In Phase 1, the stability of each state in each perceptual graph model is assessed from the point of view of the owner of the model, for each DM in the model, using standard or perceptual solution concepts, depending on the owner's awareness of others' perceptions. (In this research, only perceptual solution concepts for the 2-decision maker case are developed. ) In Phase 2, meta-stability analysis is employed to consolidate the stability assessments of a state in all perceptual graph models and across all variants of awareness. Distinctive modes of equilibria are defined, which reflect incompatibilities in DMs' perceptions and viewpoints but nonetheless provide important insights into possible resolutions of conflict. <br /><br /> The possibility principle and perceptual stability analysis are integrative techniques that can be used as a basis for empathetically studying the interaction of emotion and reasoning in the context of strategic conflict. In general, these new techniques expand current modeling and analysis capabilities, thereby facilitating realistic, descriptive models without exacting too great a cost in modeling complexity. In particular, these two theoretical advances enhance the applicability of the Graph Model for Conflict Resolution to real-world disputes by integrating emotion and perception, common ingredients in almost all conflicts. <br /><br /> To demonstrate that the new developments are practical, two illustrative applications to real-world conflicts are presented: the US-North Korea conflict and the confrontation between Russia and Chechen Rebels. In both cases, the analysis yields new strategic insights and improved advice.

Many important applications in the real world that can be modelled as combinatorial optimization problems are actually dynamic in nature. However, research on dynamic optimization focuses on continuous optimization problems, and rarely targets combinatorial problems. Moreover, dynamic combinatorial problems, when addressed, are typically tackled within an application context. <br /><br /> In this thesis, dynamic combinatorial problems are addressed collectively by adopting an evolutionary based algorithmic approach. On the plus side, their ability to manipulate several solutions at a time, their robustness and their potential for adaptability make evolutionary algorithms a good choice for solving dynamic problems. However, their tendency to converge prematurely, the difficulty in fine-tuning their search and their lack of diversity in tracking optima that shift in dynamic environments are drawbacks in this regard. <br /><br /> Developing general methodologies to tackle these conflicting issues constitutes the main theme of this thesis. First, definitions and measures of algorithm performance are reviewed. Second, methods of benchmark generation are developed under a generalized framework. Finally, methods to improve the ability of evolutionary algorithms to efficiently track optima shifting due to environmental changes are investigated. These methods include adapting genetic parameters to population diversity and environmental changes, the use of multi-populations as an additional means to control diversity, and the incorporation of local search heuristics to fine-tune the search process efficiently. <br /><br /> The methodologies developed for algorithm enhancement and benchmark generation are used to build and test evolutionary models for dynamic versions of the travelling salesman problem and the flexible manufacturing system. Results of experimentation demonstrate that the methods are effective on both problems and hence have a great potential for other dynamic combinatorial problems as well.