Wireless signal classification plays an important role in the security of a wireless communication system since it can be used for transmitter authorization, the process by which authorized transmitters are distinguished from non-authorized transmitters based on transmitter-specific traits in their signals. Current systems have only investigated authorization for a closed set scenario, where the algorithm classifies among a finite set of known transmitters. This has several limitations, the most significant of which is that transmitters outside of the known set may be misclassified. In this project, we attempt to remedy this issue by performing authentication in an open set scenario, where the number of transmitters is not known. To do this, we generated and transmitted signals from eleven different ADALM Pluto Software Defined Radios using MATLAB software. We also simulated 5 unique transmitters by artificially adding different I/Q imbalance impairments to the signals. The signals were transmitted in the form of packets (active transmission separated by idle moments), which were then extracted using pre-written code. Finally, the extracted packets were inputted into an existing deep learning algorithm called One Vs. All, where the algorithm was tested and modified until a satisfactory accuracy was found. With this algorithm, wireless communication security can be significantly improved, since it minimizes the risk of misclassification by solving the previous weakness of authorization only under closed set conditions.

How sensory information is transformed into specific ethologically appropriate motor programs is a central question in neuroscience. Precise synaptic connectivity between sensory and motor circuits is critical to preserve spatiotemporal integrity of sensory information, accurately represent the environment and produce an adaptive behavioral response. We possess limited knowledge about genetic developmental programs that regulate wiring specificity in sensorimotor circuits. To elucidate this question, we focus on visual projection neurons (VPNs) in Drosophila that serve feature detectors converting specific visual patterns into distinct motor programs. Preliminary RNAseq experiments reveal that VPN types sharing remarkably similar transcriptional programs often display striking divergence in both pre- and postsynaptic connectivity, making them attractive cellular targets to investigate molecular origins of synaptic specificity. We will primarily assess the role of differentially expressed transcription factors in regulating wiring specificity through gain and loss of function studies followed by phenotypic analyses and functional studies. These will be complemented by single cell profiling of genetically perturbed VPNs to identify specific targets of VPN-specific transcriptional programs. Gain and loss of function studies of cell surface proteins and combinations of them will be tested for direct roles in regulating synaptic specificity. Our approach will thereby link distinctive developmental programs with precise neural circuit architecture. Taken together, these studies will broaden understanding of genetic and molecular mechanisms that merge sensory and motor entities into a unified framework.

Carbon dioxide is the most abundant greenhouse gas in the atmosphere emitted through human activities. However, through an electrochemical process CO2 can be converted into more useful products such as fuels and feedstock chemicals. To improve this process so that selectivity is increased towards these more desirable products, this study will explore how mass transport affects the product distribution of the electroreduction of CO2. We alter the mass transportation of CO2 to the catalytic surface by rotating the catalyst. A cylindrical Cu catalyst is used and rotation speeds of 200 and 400 rpm are tested. The applied potential is also varied from -1.31 V vs SHE to -1.67 V vs SHE. This study provides new insights into what may be the optimal rotation speed that will produce the highest selectivity towards more desirable products, in turn opening up the possibility of CO2 reduction being a new avenue of renewable ene

Human transferrin receptor 1 (hTfR1) is a type II transmembrane glycoprotein essential for cellular iron uptake and proliferation. The acquisition of iron occurs by endocytosis upon the binding of iron-loaded transferrin (Tf) and, in certain cases, of heavy chain ferritin (HFt) to hTfR1. TfR1 is overexpressed on cancer cells in which it plays a central role in cancer cell proliferation, pathology, and survival. This makes the receptor an attractive target for cancer therapy. Thus, we developed a mouse/human chimeric antibody (ch128.1/IgG1) that binds the apical domain of hTfR1 and elicits anti-tumor activity in immunosuppressed mice bearing human multiple myeloma (MM). This activity depends on the Fc region of the antibody. Structural studies concluded that HFt also binds the apical domain of hTfR1, suggesting that ch128.1/IgG1 may compete with HFt for hTfR1 binding. Thus, we aimed to determine whether HFt and ch128.1/IgG1 compete for binding to hTfR1. Preliminary ELISA studies suggest that HFt, but not Tf, partially interferes with the binding of ch128.1/IgG1 to hTfR1, consistent with competition. These studies are relevant since inhibition of HFt by ch128.1/IgG1 may be an additional anti-tumor mechanism given the relevant role that HFt plays in cellular iron uptake in cancer cells, although it may also result in iron uptake decrease in certain normal cells. Moreover, to further understand the mechanism of in vivo activity of ch128.1/IgG1, studies are currently in progress to compare the anti-tumor activity of the parental murine 128.1 antibody and ch128.1/IgG1, which engage different FcRs, in our MM mouse model.

Glaucoma is a significant eye disease that results in blindness to millions worldwide. Glaucoma is caused by optic nerve damage resulting in permanent vision loss and dysfunction in the trabecular meshwork, a tissue near the base of the cornea by the ciliary body. Dexamethasone alters the structure of the trabecular meshwork and hinders its function, suggesting that it could serve as a therapeutic for glaucoma. A bulk collagen contraction assay and a novel single-cell contraction assay were used to measure the impact of dexamethasone exposure on contractions at the trabecular meshwork. Findings to date view expression changes of contraction-related proteins after dexamethasone treatment through RNAseq. Findings up to show that dexamethasone-treated for trabecular meshwork exhibits increased cellular contraction in both aggregate and single-cell assays, and decreased expression of cadherins. Findings also show that integrins’ expression changes after treatment appear to be donor-dependent. The integrins’ expression changes align with the literature trends in the heterogeneity of dexamethasone response to patients. What is proposed for research is that Dexamethasone may lead to increased ocular pressure by disrupting contractility through dysregulated mechanotransduction in the trabecular meshwork.

Targets to the mammalian target of rapamycin (mTOR) and other aging-related pathways have been shown to extend life span in several model organisms. Endogenous small molecules such as α-Ketobutyric acid (α-KB), whose biochemical mechanism is not well characterized, and α-ketoglutarate (α-KG), produce significant concentration-dependent extensions in C. elegans. In conducting longevity assays using incremented concentrations of α-KB and α-KG, survival of subjects was scored beginning the first day of adulthood (L4). Significant increases in lifespan in single metabolite treatment groups were reported for 2mM α-KB, 4mM α-KB, and 8mM α-KG (P <0.0001). With regards to combined metabolite treatments, 35.5% and 27.9% increases in lifespan were recorded for 2mM α-KB+8mM α-KG and 4mM α-KB+8mM α-KG groups respectively (P <0.0001). In further experimentation, additional concentrations and combinations should be tested to confirm the optimal conditions for maximal lifespan extension. While similar results have been observed with caloric-restriction-mediated downregulation of TOR activity in C. elegans, a therapeutically realistic treatment that promotes longevity should bypass any need to restrict diet or nutrient intake; results from the combinatorial treatment of α-KB and α-KG allow for further research in this direction.

Aminergic circuits play important roles in the regulation of mood, appetite, and behavior across multiple species. Aminergic neurotransmission has been shown to be required for egg laying in female Drosophila, but the mechanisms and cell types that convey this requirement remain unknown. Our lab uses aminergic stimulation of oviduct contractions as a model behavior to test presynaptic and postsynaptic hypotheses about specific mechanisms governing the aminergic egg laying circuit. Presynaptically, we seek to determine the roles of amine release from different types of vesicles by utilizing newly made aminergic transporter mutants, and, postsynaptically, we seek to understand the roles of different aminergic receptors and the different cell types that express each. Here we show that an aminergic transporter mutant that decreases amine loading into synaptic vesicles does not show an obvious oviduct contraction phenotype in optogenetic experiments. This result suggests that aminergic facilitation of oviduct behavior may occur via asynaptic transmission or indirect synaptic transmission. Using RNA interference, we also show that knockdown of amine receptors in the oviduct epithelium does not seem to affect oviduct contractions, challenging the field’s current model where oviduct muscle contractions are thought to be triggered via the adjacent epithelium. Continued work will further narrow down which amine receptors, cell types, and indirect pathways regulate oviduct contractions so that we can build better models describing the aminergic regulation of peripheral behaviors relevant to all species.

We are using a terahertz quantum-cascade vertical external cavity surface emitting laser (QC-VECSEL) with switchable polarization for polarization difference imaging. Terahertz radiation’s non-damaging nature gives it several applications such as cancer detection, bomb detection, and drug detection. Furthermore, polarimetric imaging can show various features not shown by traditional images, such as roughness, edge details, and birefringence. To optimize imaging, we first found the focal point of the laser, meaning where the gaussian beam spot size is minimum. Next, we investigated the signal to noise ratio through a wide range of parameters. We found the parameters with the strongest signal to noise ratio and from there we could move on to imaging. The beam goes through a biconvex lens, the sample, a rotating polarizer and finally two off-axis parabolic mirrors focus it onto the detector. Then we use Stokes-Mueller formalism to conveniently model the partially polarized light. Our images suggest that polarization difference can show edge features with high contrast for samples transparent in the terahertz frequency range.

A successful user interface for controlling a group of autonomous robots must have three main characteristics. First, one must be able to gather and visualize data as well as send commands to the robotic swarm in user-friendly ways. Second, because robotic projects evolve, it must be highly modular such that data streams can be easily added or displayed differently. Lastly, because it would be useful to review data from a robot’s runtime to identify and diagnose errors, the interface must be able to replay the data. To achieve these goals, I created a web-based dashboard using the JavaScript framework React. Using it, I am able to transfer data over different protocols as well as record data, which I cannot do with alternatives like IoT dashboards. I implemented different input methods such as text boxes, sliders, joysticks, and buttons, so users have flexibility in how they control robots. Anyone can now easily visualize data from multiple robots in ways including video feeds, text feeds, live graphs, and 3D rotation visualization. A user can use this dashboard to make robot development simpler as they can easily configure robots through the web interface. My dashboard was used in conjunction with other projects from my lab to control a swarm of robots that included cars, boats, and blimps. Ultimately this project enables users to easily control a generic group of robots. In the event that new controls or interfaces are needed, the modularity of this project makes it easy to implement new functionality.

Human-computer interaction (HCI) has advanced the efficacy of a multitude of sectors such as communication and consumerism. However, there exists a gap where most HCI research is conducted to improve quality in industrial aspects rather than personal aspects. Our research extends HCI to improve quality of life by designing and implementing a hand signal response AI into a six degree of freedom (6DoF) robotic arm. We call this our hand signal actuated robotic arm, SARA. An implementation of forward kinematics (FK) and inverse kinematics (IK) in python allows the robotic arm to actuate in response to complex hand signals, made possible via our hand recognition software. This software presents a real-time object-tracking process that recognizes hand signals by finger landmark mapping. A rule classifier distinguishes different variations of raised fingers. To confirm mechanical actuation and limitations, we developed a simulator in MATLAB using a virtual robotic arm that parallels SARA. Our research ultimately produced a design that, when implemented, gives SARA the capability to react to diverse hand signals independently. Qualitative demos conducted with a variety of hand signals validated our research design and implementation. A set of thirty-two hand signals was displayed to SARA that resulted in successful actuation in accordance with the simulator. The application of this design aims to assist individuals with physical limitations, making HCI more personal. The success of implementing a hand signal response AI makes the interaction with a robotic arm intuitive, ultimately expanding the scope of HCI to enhance the human experience.