DATE:Thursday, April 28, 2022
TIME:12:00 - 1:00 pm
PLACE: Zoom Seminar. A Zoom link will be emailed to the SDI list closer to the seminar; please do not share this link.
SPEAKER: Wenjun Hu, Yale University
TITLE: Edge Infrastructure Support for Diverse IoT Scenarios
ABSTRACT:
Recent years have witnessed an explosive growth of Internet of Things (IoT) devices, which collect or generate huge amounts of data. Many emerging smart-X scenarios revolve around wirelessly connected devices interacting with the environment, providing personal assistance to the device user, or automating what used to be labor-intensive tasks. However, the sheer number of devices or the volume of data generated is only part of the story. Given diverse device capabilities and application requirements, data management and processing may take place anywhere from the device itself to a nearby edge server/cloud, or a remote cloud. This creates constantly shifting bottlenecks between communication, compute, and storage.
In this talk, I will argue that the ever-increasing heterogeneity is a game-changer and calls for new thinking on infrastructure support. To that end, I will then describe two series of unconventional efforts in my group. For connectivity support, we have been experimenting with programming the radio propagation environment. This is a paradigm shift from endpoint-centric communication optimizations and paves the way for long-term infrastructure provisioning. For compute, we have designed system and service abstractions and a unique paradigm of semantic approximation to scale edge analytics. These improve resource efficiency while reducing the amount of manual efforts by orders of magnitude. Along with other ongoing efforts, we aim to design a solution suite for the overall IoT ecosystem.
BIO:
Wenjun Hu is a computer systems researcher focusing on networking and system infrastructures. She joined the Yale faculty in 2014 after spells as a researcher at Microsoft Research Asia and a postdoctoral research associate at the University of Washington. She received all her degrees in Computer Science from the University of Cambridge. During her student career she spent about 7 months each at MIT CSAIL and Microsoft Research Cambridge, as well as undergraduate summers at IBM Zurich Research Lab and EML Research in Heidelberg, Germany. She is a recipient of a Google Faculty Research Award, the 2018 SIGCOMM Test-of-Time Award, and the 2009 IEEE Bennett Prize. She is also a co-author of a widely used tool to collect channel state information, the BigStation work that laid the foundation of Microsoft Azure for Operators, and a recent research highlight in GetMobile.
VISITOR HOST: Justine Sherry
SDI SEMINAR QUESTIONS?
Karen Lindenfelser, 86716, or visit www.pdl.cmu.edu/SDI/