College of Engineering

Theses and dissertations submitted to the College of Engineering

Items in this Collection

A health information system (HIS) is computer program for automating healthcare-related tasks which may store personal information. Due to the increasing demand for data sharing between HIS, organizations implement a Health Information Exchange (HIE), which is a larger system for exchanging healthcare data among institutions, providers, and repositories. One approach in implementing HIEs is through the adoption of the Service-Oriented Architecture (SOA), a design paradigm which when followed, will provide several benefits such as increased return of investment, increased organizational agility, and reduced IT burden.

One implementation of a SOA-based HIE is the Open Health Information Exchange (OpenHIE). OpenHIE has been proven to work in several countries, however, several security features such as fine-grained access control, security levels in data, and consent management are not implemented. These features can be satisfied through the implementation of an authorization scheme using Ciphertext-Policy Attribute-Based Encryption (CP-ABE). CP-ABE is a mechanism for hiding personal information by encrypting data using attributes such as administrative role, privileges, personal profile, etc.

This work evaluated the viability of CP-ABE as an authorization scheme for SOA-based HIEs such as OpenHIE. Results show that the adoption of CP-ABE will require alterations to the HIE processes that carry out healthcare-related tasks. The results also show that the adoption of CP-ABE will require additional resources to deploy the HIE. Resources include network bandwidth, storage space in servers, and number of machines/processors. More attributes and more users lead to more resources required to finish a transaction in the HIE.

Thin film solid electrolytes have been considered as one of the solutions in decreasing the contribution of ohmic resistance in a solid oxide electrochemical cell. The commonly used oxide-ion conducting solid electrolyte is 8 mol% Yttria-stabilized Zirconia (Y0.16Zr0.92O1.92) or 8YSZ due to its low cost, good chemical stability, and good ionic conductivity. However, fabrication of dense YSZ film via facile methods on porous electrode support is still a challenge. In this study, dense 8YSZ solid electrolyte film were fabricated on a porous Lanthanum Strontium Manganite-Yttria stabilized Zirconia (LSM-YSZ) composite pellet through electrophoretic deposition (EPD). Film thickness dependence was observed by varying the deposition time. Effect of sintering temperature (1200 °C and 1300 °C) and precursor powder treatment on the film’s morphology and density were also reported. From SEM microstructural results, films with thickness of 1.5 to 9 µm were deposited on LSM-YSZ substrates. Precursor powder treatment such as ball milling and addition of dispersant reduced the particle size of the powders and subsequently produced denser films. In addition, the required sintering temperature to produce a dense film is reduced which suggests that sinterability of the films were improved by powder treatment. Morphology of the YSZ films from treated precursor powders and sintered at 1300 °C were comparable to 1500 °C-sintered bulk YSZ from untreated powders. Among the parameters used in this study, the optimal conditions to produce a dense YSZ film is by using ball milled YSZ powder with the addition of a dispersant and sintered at 1300 °C. However, further investigation is needed to produce a 100% crack-free deposit using EPD.

The analysis of networks is a significant endeavor that enables researchers to understand certain patterns and behavior, which benefits individuals and society as a whole. However, this is accompanied by privacy risks among participating individuals. Mechanisms to preserve privacy in databases have been proposed throughout the years. One of them is differential privacy, a perturbation technique wherein noise is added to the aggregated query answers to protect individuals and hide their participation in a database. In the context of networks, differential privacy is still young. It roughly falls into two directions: the release of differentially private graph statistics (i.e., the use of direct approach) and the release of synthetic graphs with differentially private parameters (i.e., the use of model-based approach). In this study, we see how each technique provides utility and privacy, and which network generation models are appropriate for different datasets and queries.

Under the context of distribution systems, a state estimator which utilizes both SCADA and µPMU measurements is considered the most beneficial form of state estimation. Individually, hybrid and multistage state estimators are only able to address few out of the many issues of state estimation. The configuration used in this paper utilizes both SCADA and µPMU measurements to both be used by a Weighted Least Squares (WLS) and a Weighted Least Absolute Values (WLAV) Estimation for the first stage and then combined using Multisensor Data Fusion (MDF) on the second stage.
Results of the WLS estimator presented in this paper show that a limited number of µPMUs on a distribution network on top of SCADA can improve the overall performance of distribution system state estimation by giving a more accurate, and reliable result. The huge difference of sampling rates of each measurement is also resolved by utilizing only one sample of each per estimation.
The results of the WLAV estimator presented in this paper show that the WLAV state estimator gave more accurate and more robust results compared to the WLS state estimator when the system is under a False Data Injection (FDI) attack. FDI attacks are used to test the robustness of the WLAV as comapred to the WLS estimator. Results show that, given the right information, the attacker can do this type of analysis first and then choose the right combination of measurements to attack in order to ensure an impact on the overall estimate. Nevertheless, even the worst results of the WLAV state estimator under multiple FDI attacks are still better than the result of the WLS state estimator under a single FDI attack. An FDI attack on a µPMU makes the WLS state estimator diverge whereas the WLAV detects it with some affected nearby buses. Countermeasures were tested on the WLAV and the recommended choice is to replace the attacked measurement by a pseudomeasurement. A linear programming formulation of WLAV (LP-WLAV) was used on the Fusion estimator because of its faster processing time than the previous WLAV formulation. The Fusion state estimator has shown more accurate and robust estimate compared to its parts – the WLS and LP-WLAV state estimators. This configuration also adds robustness against bad data and FDI attacks to the overall estimate. Dynamic state estimation is also possible under this configuration because of the fast processing time of the LP-WLAV and Fusion state estimator. This state estimator configuration which provides robustness against bad data and sub-second results is a promising solution when a limited number of µPMUs are introduced in a distribution system.

To ensure that mangrove forest conservation efforts are successful, simulation models of mangrove forests are developed to forecast outcomes of different environmental scenarios. This research presents MaDS (Mangrove species Dominance Simulator), an individual-based model that simulates the structures of mixed mangrove forest stands by considering specific responses of mangrove species to shading, salinity, and inundation frequencies. By simulating certain inundation and salinity conditions in a 50 m x 50 m plot, the model forecasts the resulting dominance of different mangrove species.
The model uses a modified Field of Neighborhood (FON) approach where species-specific shading tolerances are considered in the computation of competition response. The salinity response considers the species-specific upper-boundary salinity value of optimum growth and the species-specific maximum porewater salinity value. The inundation response considers that the limit inundation frequency that a mangrove tree can tolerate is a function of the species-specific maximum inundation frequency and the current salinity condition it senses.
Simulations of different test sites in Katunggan It-Ibajay (KII) Eco-park, Aklan, Philippines given actual site salinity and inundation frequency values showed simulated dominant species matched with dominant species in the sites given that different groups of species are tested. Simulation of mixed mangrove forests stands showed forest dynamics observed in mangrove forests, such as lack of understory trees, rise of new dominating trees in forests gaps, and the total forest above-ground biomass reaching a limit. Simulation results also showed the different abundances of species for different combinations of salinity and inundation frequency values. Results of the simulations imply that observed zonation in mangrove forests is not only caused by the environmental conditions in the site but also by the number and characteristics of species in the site.