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

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

Silicon Oxide/Oxynitride (SiO2/SiON) has been the mainstream material used for gate dielectric for MOS transistors for the past 30 years. The aggressive scaling of the feature size of MOS transistor has limited the ability of SiO2/SiON to work effectively as the gate dielectric to modulate the conduction of current of MOS transistors due to excess leakage current dominated by direct quantum tunneling. Due to this constraint, alternative gate dielectric/high-k is being employed to reduce the leakage current in order to maintain the rate of scaling of MOS transistors. However, the cost involved in the implementation of these new gate dielectric materials are high due to the requirements of a change in the process flow for device fabrication. This work presents the results of a novel processing method implementing the use of rapid thermal processing (RTP) on conventional SiO2/SiON gate dielectric to reduce the gate leakage current by three to five orders of magnitude. Electrical properties of the effect were characterized on fabricated MOS capacitors using semiconductor parameter analyzer and LCR meter. Material characterization was performed using FT-IR to understand the mechanism involved in this novel processing method, named PECE (Phonon-Energy-Coupling-Enhancement). By implementing this novel process, the use of SiO2/SiON as gate dielectric can be scaled further in conventional process flow of device fabrication.

This thesis presents numerical and analytical investigation of the geomechanics underlying longwall mining. It was tried out to study the disturbances induced by longwall mining in nearby rocks and their influence on the stability of the gates, pillars and main tunnels of longwall mines. The thesis consists of two major parts: numerical and analytical investigations. The study site is the Svea Nord coalmine, Svalbard, Norway. A novel algorithm was proposed for numerical simulation of the longwall mining process. In the proposed algorithm progressive cave-in and fracturing of the roof strata, consolidation of the cave-in materials and stress changes are simulated in detail. In order to outline the caved-in roof rocks a criterion based on maximum principal strain (in tension) was used. The critical tensile strain of roof cave-in was determined through back-calculation of the surface subsidence above a longwall panel at the mine. The results of the simulations were then used to analyse stress changes induced by longwall mining and the stability of gates. The simulations revealed that the stability of the gates and the loading to the rock bolts are closely related to the width of the chain pillars. With slender pillars, shear displacements along weak interlayers and bedding planes result in heavy loading to the rock bolts. Therefore, the locations of weakness zones should be taken into account in rock bolt design. The developed algorithm was implemented to study the loading and stability of the barrier pillar of the mine. The barrier pillars protect the main tunnels and border area of the mine from disturbances induced by longwall mining in the panels. The simulations show that the stresses in the barrier pillars fluctuate up and down during mining because of periodic cave-in events behind the longwall face. A failure zone of about 12 m exists in the wall of the barrier pillars. A large portion of the barrier pillar is still intact and is, thus, capable of protecting the border area. The results of the detailed simulations of longwall mining via the developed algorithm were, also, implemented in a large-scale numerical model. The model consists of all of the longwall panels and the border area of the mine. It is intended that the coal in the border area on the other side of the longwall panels will be mined after completion of the longwall mining. There is concern about how the longwall mining affects the stress state in the border area and how stress changes would affect future mining in the border area. A failure zone of about 20 m developed in the wall of the main tunnels on the side of the border area after all the longwall panels were mined out. The stress state in the remaining portion of the border area remains unchanged. Therefore, it will be possible to mine the border area in the future. In order to investigate the roof strata cave-in mechanism in detail a discontinuous numerical simulation of roof cave-in process was conducted by UDEC code. The block size in the roof strata and the mechanical parameters of the discontinuities were obtained through back-calculations. The back-calculations were conducted with a statistical method, Design of Experiment (DOE). Numerical simulations revealed that jointed voussoir beams formed in the roof strata before the first cave-in. Beam bending results in stress fluctuations in the roof strata. The maximum deflection of a roof stratum at the study site before the first cave-in is about 70% of the stratum thickness. The simulations and field measurements show no periodic weighting on the longwall shields in this mine. Numerical sensitivity analyses show, however, that periodic weighting may occur in strong roof strata. Roof strata with a high Young’s modulus and large joint spacing are not suitable for longwall mining. The maximum sustainable deflection of the roof strata before cave-in depends upon the horizontal in-situ stress state. It slightly increases with the in-situ horizontal stress in the stratum beams, but the horizontal stress would increase the possibility of rock-crushing in deflected roof beams. The implemented numerical method would be useful in assessment of the cavability of the roof strata and in selection of longwall shields with adequate load capacity. As shown through discontinuous numerical simulations, the roof strata above the underground opening constructed in the stratified rocks form voussoir beams. The stability of those beams is the major concern in the study of the gate stability and roof cave-in assessment in the longwall panels. Two different analytical methods were developed for cases with and without the in-situ horizontal stress acting along the beams. In the analytical model for the beams without horizontal stress a bilinear shape was assumed for the compression arch generated within the voussoir beams. The stability of the compression arch is governed by the energy method. The model requires an iterative procedure for convergence, and an algorithm was proposed for it. The analytical method was verified with numerical simulations by means of a discrete element code, UDEC. For the beams subjected to in-situ horizontal stress, the classic beam theory was employed to drive the analytical solution for it. The superposition method was used to obtain bending/deflection equations of the beam. The validity of both the assumptions and the developed method were, also, investigated by numerical simulations. The developed analytical method revealed that high Young’s modulus of a beam rock increases the stability of the beams against buckling but it causes higher stress within the compression arch which increases the probability of crushing failures in the beam abutments and midspan. In-situ horizontal stress along beams increases their stability against buckling and abutment sliding failure, but it raises the possibility of crushing failure at the abutments and the midspan.

A Field-Programmable Gate Array (FPGA) is a configurable platform for implementing a variety of logic circuits. It implements a circuit by the means of logic elements, usually Lookup Tables, connected by a programmable routing network. To utilize an FPGA effectively Computer Aided Design (CAD) tools have been developed. These tools implement circuits by using a traditional CAD flow, where the circuit is analyzed, synthesized, technology mapped, and finally placed and routed on the FPGA fabric. This flow, while generally effective, can produce sub-optimal results because once a stage of the flow is completed it is not revisited. This problem is addressed by an enhanced flow known Physical Synthesis, which consists of a set of iterations of the traditional flow with one key difference: the result of each iteration directly affects the result of the following iteration. An optimization can therefore be evaluated and then adjusted as needed in the following iterations, resulting in an overall better implementation. This CAD flow is challenging to work with because for a given FPGA researchers require access to each stage of the flow in an iterative fashion. This is particularly challenging when targeting modern commercial FPGAs, which are far more complex than a simple Lookup Table and Flip-Flop model generally used by the academic community. This dissertation describes a unified framework, called the Physical Synthesis Toolkit (PST), for research and development of optimizations for modern FPGA devices. PST provides access to modern FPGA devices and CAD tool flow to facilitate research. At the same time the amount of effort required to adapt the framework to a new FPGA device is kept to a minimum. To demonstrate that PST is an effective research platform, this dissertation describes optimization and modeling techniques that were implemented inside of it. The optimizations include: an area reduction technique for XOR-based logic circuits implemented on a 4-LUT based FPGA (25.3% area reduction), and a dynamic power reduction technique that reduces glitches in a circuit implemented on an Altera Stratix II FPGA (7% dynamic power reduction). The modeling technique is a novel toggle rate estimation approach based on the XOR-based decomposition, which reduces the estimate error by 37% as compared to the latest release of the Altera Quartus II CAD tool.

<p>Chronic pain is a distressing, prevalent, and expensive condition that is not well understood and difficult to treat. Spinal cord stimulation (SCS) has emerged as a viable means of managing chronic pain when conventional therapies are ineffective, but the efficacy of SCS has improved little since its inception. The mechanisms underlying SCS, in particular the neuronal responses to SCS, are not well understood, and prior efforts to optimize SCS have focused on electrode design and spatial selectivity without considering how the temporal aspects of SCS (stimulation frequency, pattern) may affect neuronal responses to stimulation. The lack of a biophysical basis in prior attempts to optimize therapy may have contributed to the plateau in the clinical efficacy of SCS over time. This dissertation combines computational modeling and in vivo electrophysiological approaches to investigate the effects of SCS on sensory neuron activity in the dorsal horn and uses the insights gained from these experiments to design novel temporal patterns for SCS that may be more effective than conventional therapy.</p><p>To study the mechanisms underlying SCS, we constructed a biophysically-based network model of the dorsal horn circuit consisting of interconnected dorsal horn interneurons and a wide dynamic range (WDR) projection neuron and representations of both local and surround receptive field inhibition. We validated the network model by reproducing cellular and network responses relevant to pain processing including wind-up, A-fiber mediated inhibition, and surround receptive field inhibition. To quantify experimentally the responses of spinal sensory projection neurons to SCS, we recorded the responses of antidromically identified sensory neurons in the lumbar spinal cord during 1-150 Hz SCS in both healthy rats and neuropathic rats following chronic constriction injury (CCI). In a subset of rats, we additionally assessed the impact of GABAergic inhibition on spinal neuron responses to SCS by conducting SCS experiments following the intrathecal administration of bicuculline, a GABAA receptor antagonist, and CGP 35348, a GABAB receptor antagonist. Finally, we used the computational model to design non-regular temporal patterns capable of inhibiting sensory neuron activity more effectively than conventional SCS and at lower equivalent stimulation frequencies than clinical standard 50 Hz SCS, and we experimentally validated model predictions of the improved efficacy of select patterns against conventional SCS.</p><p>Computational modeling revealed that the response of spinal sensory neurons to SCS depends on the SCS frequency; SCS frequencies of 30-100 Hz maximally inhibited the model WDR neuron consistent with clinical reports, while frequencies under 30 Hz and over 100 Hz excited the model WDR neuron. SCS-mediated inhibition was also dependent on GABAergic inhibition in the spinal cord: reducing the influence GABAergic interneurons by weakening their inputs or their connections to the model WDR neuron reduced the range of optimal SCS frequencies and changed the frequency at which SCS had a maximal effect. Experimentally, we observed that the relationship between SCS frequency and projection neuron activity predicted by the Gate Control circuit described a subset of observed SCS-frequency dependent responses but was insufficient to account for the heterogeneous responses measured experimentally. In addition, intrathecal administration of bicuculline, a GABAA receptor antagonist, increased spontaneous and evoked activity in projection neurons, enhanced excitatory responses to SCS, and reduced inhibitory responses to SCS, consistent with model predictions. Finally, computational modeling of dual frequency SCS, implemented by delivering two distinct frequencies simultaneously to distinct fiber populations, revealed frequency pairs that were more effective at inhibiting sensory neuron activity than equivalent conventional SCS and at lower average frequencies than clinically employed 50 Hz SCS. Experimental assessments of the effect of dual frequency SCS on spinal sensory neurons confirmed model predictions of greater efficacy at lower equivalent stimulation frequencies and suggest the use of non-regular temporal patterns as a novel approach to optimizing SCS. The outcomes of this dissertation are an improved understanding of the mechanisms underlying SCS, computational and experimental tools with which to continue the development and improvement of SCS. The insights and knowledge gained from the work described in this dissertation may result in translational applications that significantly improve the therapeutic outcomes of SCS and the quality of life of individuals affected by chronic pain.</p> / Dissertation

Master of Science / Department of Electrical Engineering / Dwight Day / Power consumption has become a large concern in many systems including portable electronics and supercomputers. Creating efficient hardware that can do more computation with less power is highly desirable. This project proposes a possible avenue to complete this goal by hardware accelerating a conjugate gradient solve using a Field Programmable Gate Array (FPGA). This method uses three basic operations frequently: dot product, weighted vector addition, and sparse matrix vector multiply. Each operation was accelerated on the FPGA. A power monitor was also implemented to measure the power consumption of the FPGA during each operation with several different implementations. Results showed that a decrease in time can be achieved with the dot product being hardware accelerated in relation to a software only approach. However, the more memory intensive operations were slowed using the current architecture for hardware acceleration.

Gegenstand der Arbeit stellt eine erstmalig unternommene, architekturübergreifende Studie über feldprogrammierbare Logikbausteine zur Implementierung synchroner Schaltkreise dar. Zunächst wird ein Modell für allgemeine feldprogrammiebare Architekturen basierend auf periodischen Graphen definiert. Schließlich werden Bewertungsmaße für Architekturen und Schaltkreislayouts angegeben zur Charakterisierung struktureller Eigenschaften hinsichtlich des Verhaltens in Chipflächenverbrauch und Signalverzögerung. Ferner wird ein generisches Layout-Werkzeug entwickelt, das für beliebige Architekturen und Schaltkreise Implementierungen berechnen und bewerten kann. Abschließend werden neun ressourcenminimalistische Architekturen mit Maschen- und mit Inselstruktur einander gegenübergestellt. / This work presents a first architecture-spreading study on field-programmable logical devices leaving the beaten tracks of commercial architecture improvements. After a formal model for general field-programmable architectures based on periodic graphs has been given, some feasible evaluation metrics for architectures and circuit layouts are defined characterizing structural properties of architectures in respect of chip area usage and performance. Then, a generic layout tool is developped working on arbitrary architecures and circuits. Finally, nine resource minimal mesh- and island-style architectures are compared.

This report is a preliminary exploration of the issues concerning traditional media and weblogs at two points in time. It examines the difference between the American mainstream media’s utilisation of weblogs during the 9/11 terrorist attacks in 2001, and the utilisation of weblogs in the hurricane Katrina disaster in 2005. The mainstream media’s response to the weblog form during the attacks in September 2001 and the hurricane Katrina disaster in 2005 are examined, together with an exploration of the traditional coverage. It reveals how weblogs have become a more significant part of disaster coverage by mainstream media in the United States. These processes and the implications of these changes are explored for an understanding of how the weblog is altering our understanding of the construction and dissemination of news for the traditional media institution in the United States.

Insulated Gate Bipolar Transistors (IGBTs) are popular power electronic switching devices with several advantages. However, they have been known to fail in the field when subjected to significant variations in power dissipation – known as power cycling. In the work presented here, a novel alternating-current (AC) power cycling test method for IGBTs together with their free-wheeling diodes is proposed and verified. A review of previous work revealed that the parameter that most affects IGBT lifetime under power cycling conditions is the variation in its junction-case temperature difference. Through simulation, the behaviour of a conventional single phase inverter (H-bridge) using simple pulse width modulation (PWM) control was quantified, and the effect of switching frequency and load power factor was studied. Results of the simulations and literature review were used to develop design criteria for a new AC test circuit. The new AC test circuit (a modified version of the conventional H-bridge) was then designed and its performance compared to the criteria and to the simulation results of the conventional circuit. The circuit was then built and its performance was validated. The circuit complied with the performance criteria, in particular the desired variation in 7jc, to an adequate degree of accuracy. The proposed test circuit is novel for several reasons. The stresses on devices used in a conventional H-bridge using a high power factor inductive load are reproduced using a low power factor inductive load, considerably reducing the energy cost of running such a test. IGBT switching losses are not actively reduced, as is normal practice, but instead are actively increased to generate the required losses. Free-wheeling diodes are also tested, but do not have significant switching losses, as the nature of the test circuit dictates that these be transferred to the IGBTs. The main drawback of the proposed test circuit is that a larger number of devices are needed; however, this tradeoff is necessary to obtain the energy cost savings provided by this circuit.