Scholar
Ang Chen
Google Scholar ID: 8Y4dDxkAAAAJ
University of Michigan
Citations & Impact
Citations
2,175
H-index
27
i10-index
50
Publications
20
Co-authors
50
list available
Contact
No contact links provided.
Publications
Resume (English only)
Academic Achievements
- Op-Ed on digital transformation to appear in CACM
- Vision paper on AIOps-based cloud management
- Curated and released public IaC dataset and benchmark
- Awarded $3M NSF Large project grant
- Work highlighted by NVIDIA at HotChips'23
- Published multiple papers on systems and networking, covering Poise, NetWarden, Clara, Spidermon, etc.
Research Experience
- Associate Professor, Computer Science and Engineering, University of Michigan
- Leading multiple research projects, including:
- - Digital transformation: Building a computing stack to manage large infrastructures (e.g., datacenters, power grids, water systems) and their nexuses for resilience
- - Cloud management: Advocating an AIOps-based declarative (Infrastructure-as-Code/IaC) approach; developed IaC dataset, benchmark, check generators, program lifting, and debugging tools
- - Runtime programmable networks: Enabling end-to-end, lossless, strongly consistent runtime reprogramming across host kernels, NICs, and switches; implemented runtime reconfigurable silicon switches, SmartNIC optimizations, and a program synthesis tool; awarded a $3M NSF Large project
- - Programmable in-network security: Transforming programmable networks into 'programmable defense infrastructures' with dynamic defense deployment; projects include Poise, NetWarden, Ripple, P4wn, Bedrock, RDMI, NetShuffle, SpotProxy
- - ML for systems software: Combining symbolic logic with learning-derived policies for reconfigurable low-level systems; project Clara
- - Causality in distributed systems: Using data provenance for automated fault diagnosis and prevention; projects include Spidermon, CloudCanary, Zeno, DiffProv, SPP, MetaProv
- - Infrastructure optimizations for data-intensive systems: Whole-stack co-design from network to OS to distributed frameworks for performance gains
