About
I work on networking and data-center systems—congestion control, ECN/AQM, load balancing, routing/TE—and build measurement-driven studies and event-driven simulators (ns-3/OMNeT++/SimPy/Python). My current focus is the AI × Networking seam: understanding communication patterns in distributed / decentralized training (DDP, micro-batching, tail latency) and how DCN transport and scheduling shape end-to-end performance. I completed my Ph.D. at UMass Amherst (ACQUIRE Lab, honored and fortunate to have been advised by Prof. Don Towsley during my Ph.D.).
Research Highlights
Networking & Data-Center Systems
Transport and traffic engineering; measurement and topology studies.
- Congestion control (TCP/DCTCP/CUBIC concepts), ECN/AQM (RED/CoDel).
- Load balancing (ECMP/flowlets) and routing/TE in Clos/BCube/leaf–spine.
- Topology studies; path diversity vs. blocking; measurement-driven evaluation.
Networking for Distributed Training
AI × Networking for LLM-style workloads.
- Characterizing comm patterns in distributed/decentralized training (ongoing work).
- Benchmarking DCN topologies for AI-oriented workloads and step-time sensitivity.
Distributed Quantum Computing in Data Centers
- Orchestration & windowed partitioning for remote-op heavy circuits.
- Benchmarking architectures (Clos/BCube/QFly) and fabric-contention effects.
Quantum Networking
Resource allocation and protocol analysis for entanglement distribution.
- Planning with repeater placement; distillation/swap scheduling; storage constraints.
Experience
Cisco — Research Scientist (Networking for Distributed Systems)
2025–Present
- Topology-aware DCN benchmarking and event-driven simulation for communication-bound workloads (methods generalize to LLM training).
- Scheduler-centric performance analysis; reproducible parameter sweeps and reports.
ByteDance — Research Intern (Traffic Engineering)
2021
- Prototyped RL-based traffic engineering to improve link utilization vs. an ECMP baseline in Python simulation; evaluated tail FCT and fairness.
Cisco — Research Intern (Quantum Networking)
2024
- Performance analysis of quantum protocols using a SimPy event-driven simulator; parameter sweeps and standardized figures.
Publications
-
Towards Logically Centralized Interdomain Routing
Shahrooz Pouryousef, Lixin Gao, Arun Venkataramani — USENIX NSDI (2020) -
Do LLMs Understand Collaborative Signals? Diagnosis and Repair
Shahrooz Pouryousef et al. — CIKM (2025). arXiv - Resource Allocation in Quantum Networks — PhD Thesis, UMass Amherst, 2025.
-
Resource Placement for Rate and Fidelity Maximization in Quantum Networks
Shahrooz Pouryousef, Hassan Shapourian, Alireza Shabani, Ramana Kompella, Don Towsley — IEEE TQE (2024) -
Analysis of Asynchronous Protocols for Entanglement Distribution in Quantum Networks
Shahrooz Pouryousef, Hassan Shapourian, Don Towsley — QCNC (2024) -
A Quantum Overlay Network for Efficient Entanglement Distribution
Shahrooz Pouryousef, Nitish K. Panigrahy, Don Towsley — IEEE INFOCOM (2023) -
Extortion or Expansion? Costs and Consequences of ICANN’s gTLD Experiments
Shahrooz Pouryousef, Muhammad Daniyal Dar, Suleman Ahmad, Phillipa Gill, Rishab Nithyanand — PAM (2020)
Teaching
- Instructor — First-year undergrad seminar on exploring modern computing (Fall 2023)
- Instructor — Introduction to Computer Programming (CICS 110) (Fall 2024)
- Seminar — Distributed Quantum Computing (Fall 2025)
Service
Reviewer / Community
- IEEE Transactions on Networking (ToN), 2022; IEEE ICC, 2023
Leadership & Programs
- Graduate student representative (2022–2023)
- CQN SLC industry officer (2022)
- PASP program committee member