It is our pleasure to announce the penultimate colloquium of the 2023-2024 academic year. This Friday, 3:30-4:30, we are pleased to welcome Dr. Robert Kassouf-Short, of NASA Glenn Research Center. He is a Research Engineer currently serving on the High-Rate Delay Tolerant Networking project. Please see below for the title, location, and abstract.
Title:Semiring Models for Delay Tolerant Networking
Time/Location: Apr. 19 3:30-4:30, SST 219.
Abstract: At the core of NASA’s mission is the exploration and discovery of new information about our universe. From geological satellites collecting information on melting polar ice to the Voyager missions taking pictures at the edge of our solar system, we can only collect and analyze this data if it can be communicated back to us. In a very real sense, our space communications infrastructure enables NASA to complete its mission.
As we look to bring people back to the Moon as part of the ARTEMIS program, how we maintain communication with astronauts on the Moon is a top priority. Communications architectures present technical challenges even on the surface of the Earth, where much of the infrastructure is static and already constructed. When we look to the needs for Lunar communications architectures, we find that many of our underlying assumptions about communications protocols on Earth must be abandoned in the Lunar setup. On Earth, our main connections run through cables locked into place underground; whereas on the Moon, our main connections to Earth will run through relay satellites that may or may not be in view at all times. On Earth, we have access to repair connections or restart routers if architecture stops working; whereas on the Moon, the cost of updating, restarting, or replacing a router could be (literally) astronomical.
At the networking level, the challenge of maintaining connectivity across shifting link availability is known as Delay Tolerant Networking (DTN). The foundational theory of DTN has been an active point of research for just over 5 years, and it has required the development of new mathematical models for communications. In this talk, I will present our latest cohesive mathematical model which relies on a novel semiring structure to model key aspects of DTN. Semirings provide a natural home for problems in networking, and they provide sufficient computational power to be realizable through simulations. Moreover, we leverage this model to answer questions about storage needs to achieve throughput levels across time-varying networks.
