Large language models (LLMs) are powerful models with surprising capabilities for broad downstream language processing tasks. They are however very expensive to deploy, even if pretrained. With the following seed grant from Penn State Global, we will scratch the surface towards lowering this resource barrier via optimized distributed inference through a collaboration with the Indian […]
Two papers accepted to MobiHoc ’24
The first paper extends our long-term research on network (topology) tomography to machine learning, by studying the use of tomographic information in overlay-based decentralized learning: Yudi Huang, Tingyang Sun, and Ting He, Overlay-based Decentralized Federated Learning in Bandwidth-limited Networks, ACM MobiHoc, October 2024. [full version in arXiv] The second paper results from our collaboration with Prof. […]
Tutorial at SIGMETRICS’24 and Keynote at NetSoft’24 on Network Tomography
Network tomography is a fascinating field about “reverse-engineering a blackbox network using external measurements” that combines statistical inference, performance evaluation, measurement design, and intricate algorithms. After summarizing in our book Network Tomography (cambridge.org) our learnings on a relatively easier branch of “network performance tomography” which assumes the routing topology to be known, my team has […]
Celebrating graduation for Yudi and Daniel
Congratulations on Dr. Huang (left 2) and Dr-to-be Chiu (left 3)! Congratulations again to Dr. Xie (already started at ByteDance)! And welcome to our newest member Tuan (left 1)! It is certainly a productive commencement season for NSRG with three graduates (Tian, Yudi, Daniel). Best luck on all your next endeavors.
visiting Ohio State University
I am grateful to be invited for a visit and a seminar jointly for the ECE Department of OSU and the AI-EDGE Institute: Seminar: CT Scan for Your Network: Topology Inference from End-to-end Measurements | Electrical & Computer Engineering (osu.edu) Looking forward to exploring novel research opportunities at the intersection of AI and networking.
Tian Xie passed her defense
Tian successfully defended her dissertation today, titled “Resilient Cache Network Management: Algorithms, Analysis, and Experiments”, after 5 years of study at Penn State. She will join ByteDance soon (my second alumnus at ByteDance). Congratulations to Dr. Xie!
Edited book on “Artificial Intelligence for Edge Computing” published
Our new book, “Artificial Intelligence for Edge Computing”, was just published by Springer. The book collects selected seminal results in support of Edge AI—a new genre of AI where the inference, and sometimes even the training, are performed at the point of need, meaning at the network edge where the data originate. Targeting at the […]
visiting IISc (Indian Institute of Science)
I am fortunate to be part of 5 Penn State faculty members (plus Anna Marshall from Penn State Global) selected to participate in the Penn State-IISc Joint Workshop on the campus of IISc, Bangalore, India. Great interactions with a full day of research talks from 11 faculty members from both institutions. And a fun […]
Attending MobiHoc 2023@Washington DC
Yudi at our poster (poster version of the full paper he presented at the main technical session):
Paper accepted to Sigmetrics’24
The paper, Optimized Cross-Path Attacks via Adversarial Reconnaissance, combines our investigations on network tomography (particularly topology inference) and adversarial reconnaissance to study a new type of DoS attacks arising in the context of virtualized networks (e.g., network slices in 5G). The work was started during the last year of Yilei’s PhD study and completed by […]