Search for: All records

Aridification in the U.S. Southwest has led to tension about conservation and land management strategy. Strain on multi-generational agricultural livelihoods and nearly 150-year-old Colorado River water adjudication necessitates solutions from transdisciplinary partnerships. In this study, farmers and ranchers in a small San Juan River headwater community of southwestern Colorado engaged in a participatory, convergent research study prioritizing local objectives and policy. Acknowledging the historic and sometimes perceived role of academic institutions as representing urban interests, our goal was to highlight how research can support rural governance. This process involved creating community partnerships, analyzing data, and supporting results distribution to the surveyed population through social media. The survey was designed to support a local waterway management plan. Survey results showed lack of water availability and climate changes were selected by producers as most negatively affecting their operations, and many were extremely interested in agroforestry methods and drought-resistant crop species. Statistical analysis identified that satisfaction with community resources was positively correlated with scale of production, satisfaction with irrigation equipment, and familiarity with water rights. We hope to contribute our framework of a convergent, place-based research design for wider applications in other regions to uncover solutions to resource challenges. 

We tackle the atypical challenge of supporting postquantum cryptography (PQC) and its significant overhead in safety-critical vehicle-to-vehicle (V2V) communications, dealing with strict overhead and latency restrictions within the limited radio spectrum for V2V. For example, we show that the current use of spectrum to support signature verification in V2V makes it nearly impossible to adopt PQC. Accordingly, we propose a scheduling technique for message signing certificate transmissions (which we find are currently up to 93% redundant) that learns to adaptively reduce the use of radio spectrum. In combination, we design the first integration of PQC and V2V, which satisfies the above stringent constraints given the available spectrum. Specifically, we analyze the three PQ signature algorithms selected for standardization by NIST, as well as XMSS (RFC 8391), and propose a Partially Hybrid authentication protocol—a tailored fusion of classical cryptography and PQC—for use in the V2V ecosystem during the nascent transition period we outline towards fully PQ V2V. Our provably secure protocol efficiently balances security and performance, as demonstrated experimentally with software-defined radios (USRPs), commercial V2V devices, and road traffic and V2V simulators. We show our joint transmission scheduling optimization and Partially Hybrid design are scalable and reliable under realistic conditions, adding a negligible average delay (0.39 ms per message) against the current state-of-the-art. 

We develop methods to more efficiently differentiate between gravitational wave signals from binary mergers, and detector noise. We make use of the PyCBC detection pipeline to compile larger amounts of data, including signal and noise, into SNR density plots, and we modified them so that they could be easily interpreted by an image classifier. After selecting the parameters that demonstrated features in the density plots, we created a convolutional neural network to search for these patterns. We trained and tested the neural network over increasingly large and varied data sets.