DeepQueueNet: towards scalable and generalized network performance estimation with packet-level visibility
Qingqing Yang, Xi Peng, Li Chen, Libin Liu, Jingze Zhang, Hong Xu, Baochun Li, Gong Zhang
Abstract
Network simulators are an essential tool for network operators, and can assist important tasks such as capacity planning, topology design, and parameter tuning. Popular simulators are all based on discrete event simulation, and their performance does not scale with the size of modern networks. Recently, deep-learning-based techniques are introduced to solve the scalability problem, but, as we show with experiments, they have poor visibility in their simulation results, and cannot generalize to diverse scenarios. In this work, we combine scalable and generalized continuous simulation techniques with discrete event simulation to achieve high scalability, while providing packet-level visibility. We start from a solid queueing-theoretic modeling of modern networks, and carefully identify the mathematically-intractable or computationally-expensive parts, only which are then modeled using deep neural networks (DNN). Dubbed DeepQueueNet, our approach combines prior knowledge of networks, and supports arbitrary topology and device traffic management mechanisms (given sufficient training data). Our extensive experiments show that DeepQueueNet achieves near-linear speedup in the number of GPUs, and its estimation accuracy for average and 99th percentile round-trip time outperforms existing end-to-end DNN-based performance estimators in all scenarios.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
12 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- CausalSim: A Causal Framework for Unbiased Trace-Driven Simulation cites MimicNet: fast performance estimates for data center networks with machine learning
- Scalable Tail Latency Estimation for Data Center Networks cites MimicNet: fast performance estimates for data center networks with machine learning
- DONS: Fast and Affordable Discrete Event Network Simulation with Automatic Parallelization cites MimicNet: fast performance estimates for data center networks with machine learning
- Accelerating Design Space Exploration for LLM Training Systems with Multi-experiment Parallel Simulation cites MimicNet: fast performance estimates for data center networks with machine learning
- Accelerating Design Space Exploration for LLM Training Systems with Multi-experiment Parallel Simulation cites DONS: Fast and Affordable Discrete Event Network Simulation with Automatic Parallelization
- SimAI: Unifying Architecture Design and Performance Tuning for Large-Scale Large Language Model Training with Scalability and Precision cites MimicNet: fast performance estimates for data center networks with machine learning
- SimAI: Unifying Architecture Design and Performance Tuning for Large-Scale Large Language Model Training with Scalability and Precision cites DONS: Fast and Affordable Discrete Event Network Simulation with Automatic Parallelization
Cited by top-tier papers12
- SimAI: Unifying Architecture Design and Performance Tuning for Large-Scale Large Language Model Training with Scalability and PrecisionXizheng Wang, Qingxu Li, Yichi Xu, Gang Lu et al.NSDI 2025 · 82 citations
- CausalSim: A Causal Framework for Unbiased Trace-Driven SimulationAbdullah Omar Alomar, Pouya Hamadanian, Arash Nasr-Esfahany, Anish Agarwal et al.NSDI 2023 · 46 citations
- Scalable Tail Latency Estimation for Data Center NetworksKevin Zhao, Prateesh Goyal, Mohammad Alizadeh, Thomas E. AndersonNSDI 2023 · 30 citations
- DONS: Fast and Affordable Discrete Event Network Simulation with Automatic ParallelizationKaihui Gao, Li Chen, Dan Li, Vincent Liu et al.SIGCOMM 2023 · 30 citations
- Accelerating Design Space Exploration for LLM Training Systems with Multi-experiment Parallel SimulationFei Gui, Kaihui Gao, Li Chen, Dan Li et al.NSDI 2025 · 27 citations
Builds on2
- DiffTaichi: Differentiable Programming for Physical SimulationYuanming Hu, Luke Anderson, Tzu-Mao Li, Qi Sun et al.ICLR 2020 · 479 citations
- MimicNet: fast performance estimates for data center networks with machine learningQizhen Zhang, Kelvin K. W. Ng, Charles W. Kazer, Shen Yan et al.SIGCOMM 2021 · 63 citations
Related papers
- Days: Discrete-Event Network Simulation on SteroidsBaochun LiINFOCOM 2026
- RouteNet-Erlang: A Graph Neural Network for Network Performance EvaluationMiquel Ferriol-Galmés, Krzysztof Rusek, José Suárez-Varela, Shihan Xiao et al.INFOCOM 2022 · 60 citations
- xNet: Improving Expressiveness and Granularity for Network Modeling with Graph Neural NetworksMowei Wang, Linbo Hui, Yong Cui, Ru Liang et al.INFOCOM 2022 · 36 citations
- Combining Differentiable PDE Solvers and Graph Neural Networks for Fluid Flow PredictionFilipe de Avila Belbute-Peres, Thomas D. Economon, J. Zico KolterICML 2020 · 271 citations
- Scalable Deep Learning-Based Microarchitecture Simulation on GPUsSantosh Pandey, Lingda Li, Thomas Flynn, Adolfy Hoisie et al.SC 2022 · 7 citations
