Enabling ExoComputing for Human Augmentation
While the performance of mobile processors has dramatically increased over the past decade, its processing power is still far from being sufficient to run advanced analytics services that can be utilized for human intelligence augmentation on mobile platforms. Unfortunately, this performance gap between what we have and what we wish to have is expected to grow because we are facing fundamental difficulties in persistently improving performance-per-watt of mobile processors. In response to this forecast, performing computation outside of the mobile platform, namely exo-computing, has been investigated as a possible replacement for on-device computing. However, there are significant practical hurdles to making exo-computing a reality, such as the unreliability of wireless networking. In this seminar, major technical challenges in enabling exo-computing over the cellular networks will be clarified, and potential solutions to address those challenges will be discussed.
Kyunghan Lee is an associate professor in the Department of Electrical and Computer Engineering at Seoul National University, Seoul, South Korea where he is leading the Networked Computing Lab. He received his B.S., M.S., and Ph.D. degrees from the Department of Electrical Engineering at KAIST (Korea Advanced Institute of Science and Technology), Daejeon, South Ko- rea, in 2002, 2004, and 2009, respectively.
He had been with the School of Electrical and Computer Engineering at UNIST, Ulsan, South Korea from 2012 to 2020.
He received IEEE ComSoc William R. Bennett Prize in 2013 and 2016, respectively, and received the best paper award from ACM MobiSys in 2021. He is serving on the editorial board of IEEE/ACM Transactions on Networking, IEEE Transactions on Vehicular Technology, and Computer Networks. He served as General co-Chairs for ACM MobiHoc 2022. His research interests include performance-guaranteed cellular networks and pipelined networked computing.