FPGAs are notoriously hard to use largely due to the lack of programming infrastructure. Recent synthesis tools that compile OpenCL code directly to an FPGA platform are attempting to reduce this programming burden. As part of our research in this domain, we developed the Spector benchmarks. These provide a number of different applications with tunable “knobs” that can be used to change the resource usage and performance. Furthermore, we have fully compiled each of these benchmarks with different knob setting and reported the results. The benchmarks are available in a repo under an open-source license. See our FPT paper for more information. And congrats to the authors Quentin, Alric, Pingfan, and Ryan.
It would seem like one would want a “precise” solution to security, but our recently published work shows that this isn’t always the best. In particular, by reducing the complexity of the security model, you can allow the formal solvers to finish more quickly, and in many cases give you an answer where previously you wouldn’t get one at all. The latter case happens often — the tools never fail to give an answer because you are asking them too difficult a question; they can spend days or even weeks and still not be able to determine a result. Our recent publication at the International Conference On Computer Aided Design (ICCAD) demonstrates some techniques to take advantage of this complexity vs. precision tradeoff in order to analyze hardware design for security flaws. This work was co-authored with collaborators from EPFL (Andrew Becker and Paolo Ienne), and NPU (Dejun Mu), in addition to Kastner Research Group members, Wei, Armita, and Ryan.
Antonella traveled to Shanghai to present our paper “Autonomous Acoustic Trigger for Distributed Underwater Monitoring Systems” at WUWNet ’16: The 11th ACM International Conference on Underwater Networks and Systems. The paper, authored by Antonella, Ryan, and visiting undergraduates Ethan Slattery (University of California, Santa Cruz) and Andrew Hostler (California Polytechnic San Luis Obispo) earned 1st runner up for the Best Paper Award. The paper described the development of an autonomous underwater camera system for marine population monitoring, which has been used to autonomously monitor the vaquita porpoise in Mexico, and will be used in the coming months to study the Nassau grouper in the Cayman Islands and the kelp forest soundscape in San Diego.
For the past couple of years, we have been participating in the Early Research Scholars Program (ERSP). This provides first or second year undergraduates a glimpse into the life of academic research. The goal is to allow the undergraduates to begin to understand what it means to be a researcher, write papers, deal with their annoying advisors, sit in on generally boring research group meetings, and experience all of the other activities of a university research group. It is a valuable and rewarding program run by Prof. Christine Alvarado and sponsored by the National Science Foundation. The second ERSP cohort sent our group a “thank you” card of sorts. It was very much appreciated. We certainly enjoyed working with those ERSP’ers (Proud Heng, Aishika Kumar, Rene Sanchez), and we look forward to seeing them go on to do great things. Of course, we hope that they will continue to do research with us. We would be honored to continue to have them as part of our research group.
Our group was well represented at the Oceans conference this year. Antonella Wilby and Riley Yeakle traveled to Monterey, CA to present two papers. The first paper discussed the development of an acoustic triggering system for an underwater camera trap. Antonella was the first author along Engineer for Exploration summer students Ethan Slattery (an undergraduate at UC Santa Cruz), Andrew Hostler (undergraduate at Cal Poly SLO), and Ryan. Riley presented work that details how one can use ambient acoustic ocean noise to determine the relative locations of underwater vehicles. This was joint work with Perry Naughton, Curt Schurgers, and Ryan. Of course, no visit to Monterey would be complete without a visit to the aquarium.
Links: “Design of a Low-Cost and Extensible Acoustically-Triggered Camera System for Marine Population Monitoring“, “Inter-node Distance Estimation from Ambient Acoustic Noise in Mobile Underwater Sensor Arrays“
Another Summer is gone and Fall Quarter and the new academic year is upon us. But before classes started up and campus awakened from its summer slumber, we spent a few days in Mammoth Lakes relaxing, hiking, kayaking, climbing, eating, drinking, and giving some research talks. Some of the highlights this year included homemade group dinners every night, an “easy” (compared to last year) group hike in Yosemite, and of course the compelling research discussions. Some pictures from our expert photographers (Quentin, Mike, Lu, and Ryan) are below.
The Office of Naval Research held a workshop aimed to identify key research areas for offensive of defensive hardware security research, metrics to measure high assurance, and approaches to mitigate threats. And ideally these are all deployable in the short term without modifications to the hardware. This is a tall task for sure. Ryan presented our group’s research on hardware security design tools including security testing and verification, metrics based upon information theoretic measures, and the need to employ application driven research. The workshop was hosted by Simha Sethumadhavan at Columbia University. The requisite foot picture of Columbia’s campus to prove that I was there.
The PROOFS workshop is held annually alongside CHES and CRYPTO with the goal of “promoting methodologies that increase the confidence level in the security of embedded systems, especially those which contain cryptographic algorithms”. Ryan was invited to give a keynote talk on hardware security design tools. This year the workshop was held at UC Santa Barbara, which gave Ryan the opportunity to visit his old stomping grounds (he was a professor there from 2002-2007). If you are interested, you can use the Internet “Wayback Machine” and see our old research group’s website. The painting is of UCSB Campus Point. It hangs on the wall in Ryan’s house.
Hardware Trojans are tiny pieces of circuitry that hide amongst millions, or even billions of transistors. They lay dormant until some hard to detect trigger springs them into action. Then the start their malicious behaviors like draining power or leaking secret information. The are by design difficult to detect, but our recent research shows that information flow tracking is a useful way to find and eliminate them. The research was a cover feature for the August issue of IEEE Computer. The paper’s authors are Kastner Research Group (KRG) Postdoc Wei Hu, Computer Science and Technology Ph.D. candidate Baolei Mao of Northwestern Polytechnical University (and former KRG visiting graduate student), Tortuga Logic CEO Jason Oberg, and CSE Professor Ryan Kastner. Three of the four authors (Baolei is missing) are shown a couple years ago taking in the sites at Xi’an.
Michael Barrow was one of 14 students awarded the new SIGHPC/Intel fellowship. ACM’s Special Interest Group on High Performance Computing (SIGHPC), in collaboration with Intel, established the fellowship to increase the diversity of students pursuing graduate degrees in data science and computational science, including women as well as students from racial/ethnic backgrounds that have not traditionally participated in the computing field. The fellowship starts in August, but Mike will be recognized during the high performance computing community’s flagship Super Computer (SC) conference this November. Mike’s research focuses on developing high performance vision systems to aid surgeons during operations.