Computer Network
Measurement, routing, and telemetry that keep large networks predictable under stress.
Who We Are
Welcome to UMich RobustNet Group! Led by Prof. Z. Morley Mao, we focus on computer networks, ML systems security, efficient ML systems, and autonomous vehicles.
Research Focus
MLSys Security
Defenses for data, models, and runtime stacks to resist poisoning and adversarial misuse.
Efficient MLSys
Scheduling, compression, and co-design to deliver fast, cost-aware ML across edge and cloud.
Autonomous Vehicle
Perception, control, and V2X integration that keep connected mobility resilient and verifiable.
News
- 2025 Paper accepted at NSDI 2025: "Learnings from Deploying Network QoS Alignment to Application Priorities for Storage Services"
- 2025 Paper accepted at NSDI 2025: "RLBoost: Harvesting Preemptible Resources for Cost-Efficient Reinforcement Learning on LLMs"
- 2025 Paper accepted at VehicleSec 2025: "Hijacking Attacks on UAV Follow-Me Systems in Realistic Scenarios"
Our Impacts
We actively collaborate with industry partners and our research has influenced and landed in many industrial products.
Featured Videos
Routing Support for Wide Area Mobility
Prof. Mao discusses network abstractions that keep mobile systems robust when traversing global WANs.
Intelligent Network Measurement
A deep dive into data-driven diagnosis and defense ideas that motivate RobustNet’s measurement agenda.
Security Implications in AI
Highlights from the Athena kickoff talk on aligning AI systems with robust security and policy guardrails.