Doctor of Philosophy, The Ohio State University, 2011, Electrical and Computer Engineering

The demand for wide-band small antennas is steadily increasing for both civilian and military applications due to the explosive growth of wireless communications systems. Linearly polarized electrically small antennas can be generally classified as TM10 and TE10 mode antennas. For a TM10 mode antenna, the input impedance of the antenna is considerably reactive with a small real part. In contrast, the input admittance of a TE10 mode antenna is characterized by a high susceptance and a small conductance, i.e. the input impedance is almost a short. It is therefore critical to match the antenna to a receiver (or transmitter) to optimize the transfer of power in the frequency range of interest. With conventional passive matching networks, the antennas can be only matched over narrow frequency bands. However, Non-Foster matching networks composed of negative capacitors and/or inductors can in principle match the antenna over wide frequency bands because Non-Foster matching networks can overcome the gain-bandwidth restrictions derived by Bode-Fano. In this dissertation, the design, implementation, and measurement of two Non-Foster matching networks for a TM10 mode antenna and a Non-Foster loading network for a TE10 mode antenna are the topics to be discussed, which improve performance of both types of electrically small antennas over broad frequency ranges. These devices take advantage of the unique property of Non-Foster impedances, counter-clock wise rotation on the Smith chart as the frequency increases. First, a systematic methodology is introduced to design a Non-Foster matching network for an electrically small antenna. Key steps in the proposed methodology are presented to demonstrate how to realize a fabricated Non-Foster capacitor for a 3′′ electrically small monopole receiver antenna. Based on experimental results, it is verified that Non-Foster matching networks will improve both the antenna gain and the signal to noise ratio. Second, a Non-Foster matching (open full item for complete abstract)

Committee: Roberto Rojas Ph.D. (Advisor); Fernando Teixeira Ph.D. (Committee Member); Patrick Roblin Ph.D. (Committee Member) Subjects: Electrical Engineering

Master of Computer Science (M.C.S.), University of Dayton, 2022, Computer Science

Drawing is a learned art skill for some and natural talent for others. Based on the scenario in consideration, there are times when it is necessary to assess drawing skills. In this thesis, we intend to develop an algorithm that will measure the skill of drawing by matching the hand-drawn image with the original template. The techniques that are already available make use of a complicated process. Notably, computers can be trained to identify the match at a human level which will resolve the tedious and overwhelming traditional process. Image similarity involves identifying the level of similarities in an image using a reference image; computer vision applications are used. The SIFT method and Siamese Network are analyzed and implemented to measure image similarity. The results show that it is possible to measure the art skill level. Via the analysis of the features, SIFT-based implementation was able to better detect art skills.

Committee: Tam Nguyen (Advisor); Tom Ongwere (Committee Member); James Buckley (Committee Chair) Subjects: Computer Science

Master of Science (M.S.), University of Dayton, 2015, Electrical Engineering

Intrusion Detection Systems (IDS) are intelligent specialized systems designed to interpret intrusion attempts from incoming network traffic. IDSs aim at minimizing the risk of accessing unauthorized data and potential vulnerabilities in critical systems by examining every packet entering a system. Packet inspection and Pattern matchings are often computationally intensive processes and that are the most power hungry functionalities in network intrusion detection systems. This thesis presents a high throughput, low latency and low power memristor crossbar architecture for packet header and payload matching that could be used for high-speed packet classification and malware detection. The memristor crossbar systems can perform intrusion detection through a brute force approach for static contents/signatures and a state machine approach for regular expressions. A large portion of the work completed in this thesis has been published in [1-2].

Committee: Tarek Taha Dr (Advisor); Eric Balster Dr (Committee Member); Vamsy Chodavarapu Dr (Committee Member) Subjects: Computer Engineering; Electrical Engineering

Master of Science, The Ohio State University, 2011, Electrical and Computer Engineering

A method is proposed to allow dynamic load modulation with 2 degrees-of-freedom. This method allows tuning of a load to any value within a defined coverage area of the Smith chart. This active load modulation is to be utilized in the design of an adaptive power amplifier, to maintain maximal efficiency at all power levels. Two sets of varactors, each composed of two varactors in an antiseries configuration, are used to achieve the 2 degrees-of-freedom.

Committee: Patrick Roblin (Advisor); Waleed Khalil (Committee Member) Subjects: Electrical Engineering; Engineering

Doctor of Philosophy, The Ohio State University, 2010, Electrical and Computer Engineering

Emerging broadband applications with market pressures for miniaturized communication devices have encouraged the use of electrically small antennas (ESA) and highly integrated RF circuitry for high volume low cost mobile devices. This research work focuses on developing a novel scheme to design wideband electrical small antennas that incorporates active and passive loading as well as passive matching networks. Several antennas designed using the proposed design technique and built and measured to assess their performance and to validate the design methodology. Previously, the theory of Characteristic Modes (CM) has been used mostly for antennas analysis. However; in this chapter a design procedure is proposed for designing wide band (both the input impedance bandwidth and the far field pattern bandwidth) electrically small to mid size antennas using the CM in conjunction with the theory of matching networks developed by Carlin. In order to increase the antenna gain, the antenna input impedance mismatch loss needs to be minimized by carefully exciting the antenna either at one port or at multiple ports and/or load the antenna at different ports along the antenna body such that the Q factor in the desired frequency range is suitable for wideband matching network design. The excitation (feeding structure), the loading of the antenna and/or even small modifications to the antenna structure can be modeled and understood by studying the eigenvalues and their corresponding eigencurrents obtained from the CM of the antenna structure. A brief discussion of the theory of Characteristic Modes (CM) will be presented and reviewed before the proposed design scheme is introduced. The design method will be used to demonstrate CM applications to widen the frequency bandwidth of the input impedance of an electrically small Vee shape Antenna and to obtain vertically polarized Omni-directional patterns for such antenna over a wide bandwidth. A loading technique based on the CM to eith (open full item for complete abstract)

Committee: Roberto G. Rojas PhD (Advisor); Garbacz Robert PhD (Committee Member); Teixeira Fernando PhD (Committee Member) Subjects: Electrical Engineering; Engineering; Experiments

Doctor of Philosophy, The Ohio State University, 2008, Electrical and Computer Engineering

This dissertation presents several methods to more efficiently use the computational resources availableon a Heterogeneous Chip Multiprocessor (H-CMP). Using task scheduling techniques, three challenges to the effective usage of H-CMPs are addressed: the emergence of reconfigurable hardware in general purpose computing, utilization of the network on a chip (NoC), and fault tolerance. To utilize reconfigurable hardware, we introduce the Mutually Exclusive Processor Groups reconfiguration model, and an accompanying task scheduler, the Heterogeneous Earliest Finish Time with Mutually Exclusive Processor Groups (HEFT-MEG) scheduling heuristic. HEFT-MEG schedules reconfigurations using a novel back-tracking algorithm to evaluate how different reconfiguration decisions affect previously scheduled tasks. In both simulation and real execution, HEFT-MEG successfully schedules reconfiguration allowing the architecture to adapt to changing application requirements. After an analysis of IBM's Cell Processor NoC and generation of a simple stochastic model, we propose a hybrid task scheduling system using a Compile- and Run-time Scheduler (CtS and RtS) that work in concert. The CtS, Contention Aware HEFT (CA-HEFT), updates task start and finish times when scheduling to account for network contention. The RtS, the Contention Aware Dynamic Scheduler (CADS), adjusts the schedule generated by CA-HEFT to account for variation in the communication pattern and actual task finish times, using a novel dynamic block algorithm. We find that using a CtS and RtS in concert improves the performance of several application types in real execution on the Cell processor. To enhance fault tolerance, we modify the previously proposed hybrid scheduling system to accommodate variability in the processor availability. The RtS is divided into two portions, the Fault Tolerant Re-Mapper (FTRM) and the Reconfiguration and Recovery Scheduler (RRS). FTRM examines the current processor availability and remap (open full item for complete abstract)

Committee: Fusun Ozguner (Advisor); Umit Catalyurek (Committee Member); Eylem Ekici (Committee Member) Subjects: Computer Science; Electrical Engineering

Master of Science (MS), Ohio University, 1981, Electrical Engineering & Computer Science (Engineering and Technology)

Optimum design of broadband microwave transister amplifiers

Committee: Wai Chen (Advisor) Subjects:

PHD, Kent State University, 2012, College of Arts and Sciences / Department of Computer Science

With the rise of the internet, mobile communications, electronic transactions, and personal broadcasting, the scale of connectedness has grown immensely. Not only can an individual interact with thousands and millions of others, but details about those interactions are being stored in databases, for later retrieval and analysis. Two key concepts help us to simplify and understand networks: structural patterns and social role. Networks often exhibit recurring structural patterns, and similar structure often correlates to similar functional or behavioral role. The presence of recurring roles and structural patterns also enables transfer learning: what we know about one network can be used to help us understand or identify information in another network. While the theoretical concept of structural roles is well-established, however, there is no agreed-upon real-valued measure of role similarity. In this work, we focus on two specific computational problems: (1) developing a well-principled and scalable measure for node structural similarity, and (2) finding the optimal node-to-node alignment between two graphs. This dissertation makes the following contributions. First, to establish a sound theoretical basis, we present an axiomatic definition of a role similarity measure. This proves a clear and uniform understanding for the characteristics needed by any role similarity. The key axiom is automorphism/isomorphism confirmation: if two nodes are automorphically equivalent, then an admissible similarity measure must positively confirm this fact. Second, we present RoleSim, a role similarity metric which satisfies these axioms and which can be computed with a simple iterative algorithm. RoleSim is founded on the concept of maximal matching of neighbor similarity. We rigorously prove that RoleSim satisfies all the axiomatic properties and demonstrate its superior interpretive power of RoleSim both synthetic and real datasets. Third, we establish a recursive connection betw (open full item for complete abstract)

Committee: Ruoming Jin (Advisor) Subjects: Computer Science