Search

This thesis examines a Distributed Interference Impact Probing (DIIP) strategy for Wireless Ad hoc Networks (WANETs), using a novel cross-layer Minimum Impact Routing (MIR) protocol. Perfonnance is judged in tenns of interference reduction ratio, efficiency, and system and user capacity, which are calculated based on the measurement of Disturbed Nodes (DN). A large number of routing algorithms have been proposed with distinctive features aimed to overcome WANET's fundamental challenges, such as routing over a dynamic topology, scheduling broadcast signals using dynamic Media Access Control (MAC), and constraints on network scalability. However, the scalability problem ofWANET cannot simply adapt the frequency reuse mechanism designed for traditional stationary cellular networks due to the relay burden, and there is no single comprehensive algorithm proposed for it. DIIP enhances system and user capacity using a cross layer routing algorithm, MIR, using feedback from DIIP to balance transmit power in order to control hop length, which consequently changes the number of relays along the path. This maximizes the number of simultaneous transmitting nodes, and minimizes the interference impact, i.e. measured in tenns of 'disturbed nodes'. The perfonnance of MIR is examined compared with simple shortest-path routing. A WANET simulation model is configured to simulate both routing algorithms under multiple scenarios. The analysis has shown that once the transmitting range of a node changes, the total number of disturbed nodes along a path changes accordingly, hence the system and user capacity varies with interference impact variation. By carefully selecting a suitable link length, the neighbouring node density can be adjusted to reduce the total number of DN, and thereby allowing a higher spatial reuse ratio. In this case the system capacity can increase significantly as the number of nodes increases. In contrast, if the link length is chosen regardless ofthe negative impact of interference, capacity decreases. In addition, MIR diverts traffic from congested areas, such as the central part of a network or bottleneck points.

At the UC Davis Fish Conservation and Culture Laboratory (FCCL), Delta Smelt (Hypomesus transpacificus) and Longfin Smelt (Spirinchus thaleichthys) are being cultured for conservation. While the aquaculture processes for these species were established, they are not perfect and demand extensive labor and time, particularly for fish egg treatments. This study explored alternative egg separation and adhesiveness removal methods. In the first part of the study, Epsom salt solutions with varying specific gravity (SG) were utilized to separate live and dead eggs. The results showed promising effectiveness for specific gravity value of 1.08, 1.10, and 1.12, which effectively separated dead eggs for both species with minimal impacts on hatching rates. In the second part of the study, Ramshorn Snails (Helisoma anceps) were utilized as natural egg separators. The result showed these snails preferred dead Delta Smelt eggs compared to live eggs. This behavior was first discovered and opened a potential route for future fish egg separation methods. In the last part of the study, alternative fish egg adhesiveness removal methods were compared to the traditional bentonite clay method. The results showed excellent egg adhesiveness removal capability for sodium hypochlorite and tannic acid, while the effectiveness of Alcalase enzyme increased with concentration and achieve the same effectiveness at the higher concentration of 5% (v/v). However, the results also showed the negative impacts of sodium hypochlorite to the hatching rate of Delta Smelt eggs, which no eggs hatched after being treated. Tannic acid and Alcalase, on the other hand, showed no significant impact to the hatching rate. From the findings of this study, the Epsom salt solution method for fish egg separation has been successfully integrated to the hatchery-scale fish production at the FCCL, yielding notable time and labor savings for egg treatments. However, the approaches using Ramshorn Snails and alternative adhesiveness removal techniques require further study and optimization.

Limited sensing capabilities pose challenges to autonomous vehicles. Common limitations can result from a limited sensory range, occlusions, or adverse weather conditions, and hinders the development of both general and safe autonomy in real-world environments. In the pursuit of safe autonomy under imperfect measurements, an autonomous vehicle must assess the risk to take safe actions, and actively gather informative data to reduce the uncertainty. This thesis develops efficient algorithms for risk assessment and noise removal, specifically with Light Detection And Ranging (LiDAR), to aid the safe navigation of autonomous vehicles. First, a probabilistic risk assessment algorithm is proposed, which utilizes scene geometry to quantify the risk imposed by occluded oncoming traffic. Simulations show that the quantified risk can then be used as an objective function of a subsequent planning algorithm to improve both safety and ride comfort at real-world urban intersections. Second, the efficiency of the aforementioned algorithm is further enhanced by leveraging reachability analysis to constrain its sensory resources to only regions with potential collisions. This leads to an algorithm that not only quantifies the risk more efficiently, but also localizes the risk-inducing traffic participants in an environment. As a result, unlike the previous approach, the vehicle no longer needs to put equal attention to the entire scene and could focus only on essential information in collision avoidance. Lastly, a real-time de-noising algorithm is proposed to remove noise from LiDAR point clouds corrupted by snow. This reduces uncertainty from adverse weather conditions. The proposed network outperforms state-of-the-art methods while also running orders of magnitude faster. The de-noised point clouds also show positive impacts on off-the-shelf Simultaneous Localization And Mapping (SLAM) algorithms, resulting in maps nearly free of noise while preserving salient information of the scene. These algorithms are efficient and make safe autonomy feasible in the near future, even under imperfect LiDAR measurements.

A common and serious complication of radiation treatment for cervical cancer is vaginal stenosis, defined as the narrowing or shortening of the vaginal canal. Prior studies have found that vaginal dilators can be an effective early treatment for vaginal stenosis but suffer from low patient adherence due to pain and discomfort. Inflatable silicone balloon dilators have recently been designed and implemented in the Talke lab to address low patient adherence of traditional dilators. By taking advantage of the high flexibility of silicone and the continuously adjustable size of an inflatable balloon, the new dilator design reduces pain and discomfort for the patient. To simulate a stenosed vaginal canal and evaluate the efficacy of the inflatable dilator, we have built a vaginal model using a specially designed composite material consisting of silicone and thermoplastic polyurethane. This material can create a vaginal phantom with stiffness variation. The two-material composite allows us to tune the stiffness of our vaginal tissue analog, a critical symptom of vaginal stenosis. This thesis investigates this newly designed composite material of silicone and thermoplastic polyurethane to help in the understanding of the treatment of vaginal stenosis. Investigating how stenosed vaginal tissue reacts to an inflatable dilator will be a guide to clinical applications and the assessment of vaginal stenosis. Tensile coupons were manufactured and tested with a uniaxial tensile tester and a digital image correlation setup. The constitutive equations for the silicone-polyurethane composite were determined and analyzed in terms of Mooney-Rivlin coefficients for hyperelastic materials.

This thesis considers estimation and statistical inference for high dimensional model with constrained parameter space. Due to the recent development of data storage and computing technology, it is extremely common for researchers to face a high dimensional problem in practical applications, ranging from health-care, neural imaging, genetic studies, etc. In a high dimensional problem, the number of unknown parameters is usually much larger than the sample size, imposing additional difficulties on accurately estimating the parameters. As a result, it is usually assumed that the parameter satisfies some certain constraints, such as sparsity constraint or low-rank constraint. In this thesis, we develop novel algorithms to obtain accurate parameter estimation and statistical inference for several high dimensional models with constrained parameter space. Chapter 2 discusses asymptotic inference for high dimensional model under equality constraint. We propose a novel inference method that takes the equality constraint into consideration. The proposed estimator enjoys asymptotically smaller variance than the standard method without constraints, and is semiparametric efficient. Chapter 3 considers high dimensional statistical inference with inequality constraint. We develop tools to test whether the parameters are on the boundary of the constraint or not. The proposed testing procedure has greater power than the standard algorithms where the constraints are ignored. Chapter 4 studies the problem of recovery of matrices that are simultaneously low rank and row and/or column sparse. We propose a GDT (Gradient Descent with hard Thresholding) algorithm that converges linearly to a region within statistical error of an optimal solution. Chapter 5 considers the safe reinforcement learning problem. We construct a sequence of surrogate convex constrained optimization problems by replacing the nonconvex functions locally with convex quadratic functions obtained from policy gradient estimators. We prove that the solutions to these surrogate problems converge to a stationary point of the original nonconvex problem.

The development and commercialization of rechargeable Li-ion battery in the 1990s has triggered the advancement of modern portable technology. Currently, with the emergence of electric vehicles and more complex gadgets, conventional lithium ion intercalation based secondary batteries could no longer live up to the demands of the end-use consumers. In this dissertation, ordered carbonaceous nanomaterials were utilized to integrate with the next-generation conversion chemistry based secondary batteries. Lithium-oxygen and sodium-sulfur battery systems were tested and developed in attempt to deliver higher energy density over the conventional Li-ion battery. The main objective of the work is to fabricate advanced electrodes that are capable of providing higher capacity in order to facilitate power grids and electric vehicles, while by doing so; the effects caused by the emission of CO2 could be significantly mitigated. Ordered carbonaceous nanomaterials were produced by the chemical vapor deposition method. Among the wide variety of carbon materials, vertically aligned carbon nanotubes were grown on a substrate and then subsequently peeled off and used as the electrode for the lithium-oxygen battery. With the combination of 2-methyl-pyrrolidone solvent, the assembled lithium-oxygen battery could achieve specific capacity of 1200 mAh·g-1 and under safe charge/discharge cycles for 50 cycles. Carbonized metal organic framework was fabricated by mixing selected precursors. Sulfur was melt infiltrated to yield the carbon/sulfur composite. The metal organic framework structure composite cathode exhibited specific capacity of 1000 mAh·g-1 for over 250 cycles for the room temperature sodium-sulfur system. In addition, nitrogen, sulfur co-doped hierarchical porous carbon was fabricated by soft template method and combined with sulfur. The N,S-HPC/S composite showed relatively lower energy density; however, with much higher cycle stability of ~10,000 cycles at a current density of 4.6A·g-1. The next generation secondary battery showed good performance yet with limited protection for the inevitable lithium dendrite growth if coupled with metallic lithium, which is an adverse phenomenon that would eventually result in the deterioration of the battery. The shortage of the battery is caused by the dendrite penetration, which could lead to safety issues including thermal runaways and explosions. Therefore, solid polymer electrolyte was designed and prepared to mechanically suppress the growth of dendrite during charging protocol. PEGDA polymer host, plasticizers and lithium salt were used to fabricate the solid ionic conducting membrane, which could achieve an ionic conductivity of 10-3 S·cm-1, a value that is comparable to the liquid electrolyte counterpart. The solid polymer electrolyte was implemented in the conventional lithium-ion battery and the advanced lithium-sulfur battery.

This dissertation is principally concerned with the combined state and parameter estimation problem, where the goal is to estimate the state of a discrete-time, linear time-invariant system with structured uncertainty in the system dynamics. First, we prove necessary and sufficient conditions for the identifiability of unknown parameters within a state-space realization. Next, we evaluate the performance of classical techniques for solving the combined state and parameter estimation problem. We then formulate and test the retrospective cost parameter estimation algorithm under the assumption that the initial states are known. Two variants of the retrospective cost parameter estimation and smoothing algorithm are formulated and tested in the case where the initial states are unknown. Finally, the retrospective cost Kalman filter algorithm is formulated and tested for state estimation despite uncertain noise covariances and potentially nonzero-mean sensor and process noise.

Wheat is one of the most important crops in the world. Wheat can be classified according to its qualitative requirement for low temperature to accelerate flowering (i.e. vernalization). Winter wheat has a vernalization requirement whereas spring wheat does not. Winter wheat can be further classified according to its quantitative vernalization requirement: weak winter, semi-winter, and strong winter types. Three genes were cloned based on qualitative variation in vernalization requirement between spring and winter wheat, but genes controlling quantitative variation among winter wheat cultivars are not well characterized. In this study, the following genetic features for wheat development were addressed.

The opening of the 21st century has been marked as a generation of biological science. Nowadays, the understanding of the sequence and structure of biomolecules is growing rapidly. And researchers from multiple disciplines, chemistry, physics and computational science in particular, are making significant contributions to modern biology and drug discovery. The state-of-the-art techniques of computational chemistry and molecular modeling can be applied to study a wide range of chemical and biological systems of interest. This enables us to study structural details at the atomic level and obtain chemical/biological information which is not available by experimental measurements. This dissertation project focused on modeling the recognition mechanisms of biomolecules and their conjugated ligands. Multiple computational techniques, such as molecular dynamics simulation, entropy and free energy calculation were applied. The model systems included signaling domains (FHA, BRCT and WW domain), kinase (p38 kinase) and HIV protease. The details of the dynamics, interactions and correlations of domain-phosphopeptide systems were discovered. The free energy calculation, mining minima algorithm, was performed to study detailed conformational changes of kinase-ligand systems and predict protein-ligand binding energies. In addition, the association process of protein and ligand, which requires large simulation time scales, was also investigated. This project studies the details of protein-peptide/ligand binding and provides clues for further structure-based biomolecule/drug design.

Abstract: This thesis work is aimed at addressing some of the fundamental questions surrounding the intracellular fate of plasmid DNA in clinically-relevant primary cells, when delivered with lipid-modi!ed cationic polymers as gene carriers. We developed an optimized procedure for the transfection of primary cells, which included modi!cations designed to maximize the in vitro stability of the DNA-carrier complexes and enhance their transfection utility. These polymeric gene carriers bind to DNA through electrostatic interaction, which drives a self-assembly process that condenses DNA into sub-micron particles suitable for cellular uptake. This interaction is not DNA sequence or structurally speci!c, and thus able to package DNA molecules with different topologies and molecular weights with equal efficiency for uptake. However, circularized DNA performed better than its linearized equivalent in transfection, suggesting intracellular processing may be dependent on the physicochemical properties of the assembled complexes. Thus, we concentrated our effort on understanding the intracellular events leading to transfection. Using a linoleic acid substituted cationic polymer, we found that transfection efficiency is correlated with the amount of pDNA associated with the nucleus and that lipid-moieties is able to facilitate nuclear association of DNA to enhance transfection. Further, lipid modi!cation altered the uptake pathways of the complexes from ones that are predominantly driven by macropinocytosis to ones that are mediated by clathrin. Endosome escape continues to be an inefficient process with these polymeric gene carriers, suggesting methods to promote cytosolic release may be the most effective approach to enhance their transfection efficiencies.

Donor-Acceptor (DA) cyclopropanes are considered to be well-understood compounds with predictable chemistry. This dissertation focuses on our recent advances and synthetic applications of DA cyclopropanes principally involving 1,3-dipole intermediates revealed by reactions with Lewis acids. Particular emphasis is placed on reactions involving sugar-derived substrates. Chapter one is an overview of the literature preparations and synthetic applications of DA cyclopropanes. It begins with general DA cyclopropanes but is mostly focused on carbohydrate-derived cyclopropanes. Chapter two describes our intramolecular glucal cyclopropanation approach toward lactonized cyclopropanes 2.5a-c, their transformations under mild conditions, and their application to the quasi formal synthesis of the natural product xylobovide 2.34a. Chapter three investigates the different chemical behaviors of glucal-derived lactonized cyclopropane 2.5a under the influence of a variety of different Lewis acids. Use of boron trifluoride etherate (BF3•OEt2) leads to regioselective mono-desilylation of the di-tert-butylsilylene ether to 2.27; trimethylsilyl trifluoromethanesulfonate (TMSOTf) results in anhydro-sugar 3.8 formation; and titanium tetrachloride (TiC4) facilitates stereoselective allylation and glycosidation. Chapter four discuses our recently discovered formal [3+2] dipolar cycloaddition reactions of DA cyclopropanes with a wide variety of dipolarophiles including silyl enol ethers, imines, aldehydes and nitriles. The extension of nitrile cyclization to general noncarbohydrate derived DA cyclopropanes leads to the discovery of a novel synthetic methodology toward pyrroles, bipyrroles and thienylpyrroles. Finally, the [3+2] cycloaddition has also been extended to heterocycles including pyridines, quinolines and indoles. Chapter five summarizes the investigation of regio-selective mono-desilylation of di-tert-butylsilylene ethers as a new protocol for selective hydroxy protection. Under optimized conditions, the reaction demonstrates high regioselectivity as well as strong functional group compatibility, and proves promising as a new protocol for selectively manipulating 1,3-diols. Chapter six provides the experimental procedures and characterization of all the new compounds.

The AL2 protein of tomato golden mosaic virus (TGMV) is multifunctional. It is required for derepression of the TGMV AR1 gene in phloem tissue, and for trans-activating the AR1 and BR1 genes in mesophyll cells. It also enhances virus susceptibility when expressed in transgenic plants. It is thought that TGMV AL2 protein accomplishes these functions by interactions with unknown host factors. In this study, cDNAs from Nicotiana benthamiana plants that encode proteins which interact with TGMV AL2 were characterized. A yeast two-hybrid assay identified two cDNA clones, Nb#51 and Nb#62, that specifically interacted with TGMV AL2, but not with negative control proteins. Sequences of these two cDNA clones were determined by primer walking, which revealed that Nb#51 appears to be a 3'-coterminal truncated version of Nb#62. Inspection of the amino acid sequences encoded by Nb#62 found the presence of both ankyrin-repeats and tetratricopeptide repeats (TPR). Deletion analysis showed that the TPR motif, together with its flanking regions, was sufficient to confer on Nb#62 the ability to interact with TGMV AL2, whereas the ankyrin-repeats were not required for this interaction. Nb#62-specific mRNAs were detected in N. benthamiana plants by northern hybridization in potato virus X (PVX) infection experiments, but not in heat-shock experiments. Virus-induced gene silencing (VIGS) assays were used to investigate the possible function(s) of the Nb#62-encoded protein in normal plants and in the context of a TGMV infection. When PVX carrying a 618-bp fragment from the 5'-end of the Nb#62 cDNA was used as a silencing trigger in N. benthamiana plants, VIGS effectively targeted the transcripts which contained sequence similarity with the trigger fragment. However, the infected plants didn't have difference in the phenotype when compared to the PVX vector infection. When TGMV carrying a 93-bp fragment from the 5'-end of the Nb#62 cDNA infected plants, viral DNA accumulation in the upper leaves was reduced, when compared to wild-type TGMV or a control TGMV construct containing a 95-bp fragment from the tobacco Sulfur gene.

In this project, an H∞ controller is proposed as a DC voltage regulator for the predictive current controlled pulse with modulated (PWM) three-phase power converter to compensate the uncertainty caused by the predictive current controller. The H∞ theory has its unique approach to the uncertainty issue, its design method allows both the stability robustness and performance robustness to be considered at the design stage. The design of the H∞ voltage regulator is based on the worst case scenario where the converter is subject to a load disturbance and its operation mode is changed from the rectifying to the regeneration, consequently, parameters and structure of the original plant transfer function derived, vary significantly. The proposed H∞ (DC) voltage regulator is to overcome the uncertainty problem and to work with the predictive current controller to achieve robust control of the entire system. The combined control scheme is simulated using MatLab/SIMULINK environment. The simulation results show that the closed-loop system is capable of achieving performance robustness and stability robustness for the worst case scenario. The output DC link voltage is stable and nearly sinusoidal, while the line currents are delivered with a unity power factor. A computer controlled three-phase PWM converter of lOkVA is developed and tested in the open-loop condition, and the performance of the power converter is examined. An INTEL industrial PC-486 single chip computer is used as a digital controller. For a given modulation index and output frequency, switching intervals are calculated using the space vector PWM algorithm, and are down loaded to a PWM generator board which synthesizes gating pulse to six switching devices (IGBT). There is a on-board synchronizing unit which synchronizes the fundamental frequency of the gating pulses with the mains. A computer program is developed in Borland C++ to perform the modulation process. The experimental results show that the developed converter system is ready to be used as a test bed for conducting experimental research in the area of converter control and machine control. To summarize the work accomplished by the author for this project: 1. The introduction of the H∞ control theory into the application of converter control and theoretical design of the H∞ voltage regulator; 2. The simulation of the entire closed-loop controlled power converter with the combination of a predictive current control and a H∞ voltage regulator; 3. The development of a computer controlled three-phase PWM power converter as a test bed.

Zygospores of Choanephora cucurbitarum (Berkeley et Ravevel) Thaxter germinated by producing germ tubes with terminal germsporangia, germsporangioles on vesicles or germ tubes which continued to grow without producing germsporangia. Zygospores have a light requirement for germination and were stimulated by KMnO4, NaClO and H2O2. Zygospore progeny from crosses between (+) and (-) isolates of g. cucurbitarum consisted of (+), (-) and (+,-) mating types. Mating types of germsporangiospores in each germsporangium were either all (+), or all (-) from respective (+) or (-) zygospore cultures and, all (+,-), or (+) and/or (-) plus (+,-) types from (+,-) zygospore cultures. Mating type heterokaryons (+,-) obtained from zygospore and germsporangiospore cultures in all crosses produced azygospores and zygospores which segregated into (+) and (-) during asexual propagation. All the zygospore and germsporangiospore progenies from the cross between two fast growing isolates were of fast growing type, while all those from the crosses between two slow growing isolates were of slow growing type. The crosses between fast and slow growing isolates gave rise to progeny with both fast and slow growing types. Mating types of zygospore isolates from crosses between two near identical isolates followed 1(+):1(-):2(+,-) ratio. The distribution of mating types (+) and (-), and growth types F and S fit a 1:1 ratio in zygospore progeny in all crosses. Results suggest that mating type as well as growth type are each controlled by a gene with a pair of alleles, and that these two genes are linked to each other with a distance of about 33 cM. The heterokaryotic zygospore progeny from the crosses between F and S types consisted of F(+,-), S(+,-), F,S(+,-), S,F(+,-), F,S(+) and F,S(-). Each of them segregated into two types during sporangiospore formation, suggesting that only two of the four nuclei resulted from meiosis move into the germ tube of a germinating zygospore. Mating type heterokaryons (+,-) formed through fusion of (+) and (-) protoplasts. The (+,-) colonies from protoplast fusion also produced azygospores and zygospores which segregated into (+) and (-) mating types during sporangiospore formation.