Changgang Zheng

Researcher | DPhil, University of Oxford

Email: changgang [dot] zheng [at] eng [dot] oxfordalumni [dot] org
Or: changgangzheng [at] qq [dot] com

Publications  |  Group  |  Talks  |  Service  |  Teaching  |  Honors  |  Google Scholar  |  GitHub

profile photo

I was a researcher at Alibaba Cloud and a Postdoctoral Researcher at University of Oxford. I obtained my DPhil in Engineering Science from Oxford, advised by Prof. Noa Zilberman. I had the pleasure of working with Prof. Leandros Tassiulas during a research visit at Yale University. I received dual BEng degrees from the University of Glasgow (first-class honours) and UESTC.

My research interests include data center networks, AI computing networks, and space networks. I am particularly interested in leveraging in-network computing and machine learning techniques to design efficient and scalable network systems and address fundamental challenges in networking.

Selected Publications [Full List]
  • "Scaling LLM Agent Tool Access: The MCP Gateway", ACM SIGOPS ATC 2026
  • "CStar Gateway: Augmenting Public Cloud Infrastructure for Heterogeneous Network Function Virtualization", USENIX NSDI 2026 [pdf]
  • "Bifrost: Alibaba's Next-Generation VPC Network with High-Performance Multipath Reliable Transport", USENIX NSDI 2026 [pdf]
  • "IIsy: Hybrid In-Network Classification Using Programmable Switches", IEEE/ACM TON 2024 [pdf]
  • "Planter: Rapid Prototyping of In-Network Machine Learning Inference", ACM SIGCOMM CCR 2024 [Best of CCR] [pdf]
  • "In-Network Machine Learning Using Programmable Network Devices: A Survey", IEEE Communications Surveys and Tutorials 2024 [pdf]
  • "DINC: Toward Distributed In-Network Computing", ACM CoNEXT 2023 [pdf]
Selected News
MCP Gateway was accepted to ATC 2026.
The 2026 ACM SIGOPS Annual Technical Conference (ATC)
[Acceptance Rate: 135/973=13.9%] [CCF A]
(Paper to be released)
ParaFlowO and Strawman were accepted to APNet 2026.
The 10th Asia-Pacific Workshop on Networking (APNet 2026)
[Acceptance Rate: 39/147=26.5%]
ParaFlowO | Strawman | MCP Gateway Poster
CStar Gateway and Bifrost were accepted to NSDI 2026.
The 23rd USENIX Symposium on Networked Systems Design and Implementation (NSDI '26)
[Acceptance Rate: 100/452=22.1%] [CCF A]
CStar Gateway | Bifrost
Presented "MUTA: Enabling Multi-Task Neural Network Inference in Programmable Data-Planes" on the 22nd of May, 2025.
Kaiyi Zhang, Changgang Zheng, Nancy Samaan, Ahmed Karmouch, and Noa Zilberman
The 26th IEEE International Conference on High Performance Switching and Routing (HPSR 2025) [CORE B] [Best Paper Award]
Paper | Link | BibTex
Accepted "Planter: Rapid Prototyping of In-Network Machine Learning Inference", 2024
Changgang Zheng, Mingyuan Zang, Xinpeng Hong, Liam Perreault, Riyad Bensoussane, Shay Vargaftik, Yaniv Ben-Itzhak, and Noa Zilberman
ACM SIGCOMM Computer Communication Review (CCR), 2024 [Acceptance rates comparable to ACM SIGCOMM conference, i.e., under 15%] [Best of CCR]
PDF | arXiv (2022) | BibTex | Code | Doc
Accepted "IIsy: Hybrid In-Network Classification Using Programmable Switches", 2024
Changgang Zheng, Zhaoqi Xiong, Thanh T Bui, Siim Kaupmees, Riyad Bensoussane, Antoine Bernabeu, Shay Vargaftik, Yaniv Ben-Itzhak, and Noa Zilberman
IEEE/ACM Transactions on Networking, 2024 [CCF A]
PDF | arXiv (2022) | BibTex | Code
Accepted "In-Network Machine Learning Using Programmable Network Devices: A Survey" on the 7th of December, 2023.
Changgang Zheng, Xinpeng Hong, Damu Ding, Shay Vargaftik, Yaniv Ben-Itzhak, and Noa Zilberman
IEEE Communications Surveys and Tutorials [Impact Factor=35.6]
Paper | Link | BibTex
Accepted "DINC: Toward Distributed In-Network Computing"
Changgang Zheng, Haoyue Tang, Mingyuan Zang, Xinpeng Hong, Aosong Feng, Leandros Tassiulas, and Noa Zilberman
ACM CoNEXT'23 & Proceedings of the ACM on Networking, 2023
[Acceptance Rate: 24/129=18.6%] Paper | BibTex | Code
More ...