NeuralEmu: in situ Measurement-Driven, ML-based, High-Fidelity 5G Network Emulation

2026-04-28

AI Summary

Existing 5G network simulators struggle to accurately reproduce the dynamic feedback between commercial base station schedulers and applications, leading to significant discrepancies between simulated and real-world behavior. This work proposes a machine learning–based, high-fidelity 5G simulation framework that leverages high-resolution network telemetry data to learn base station scheduling policies, dynamically predicting resource block allocation and modulation schemes. It further incorporates a traffic inversion model to reconstruct background user traffic, enabling, for the first time, multi-client scheduler behavior emulation. Implemented as a high-performance Linux middlebox, the framework reduces simulation errors by 55%, 57%, and 51% for key performance indicators—web page load time, WebRTC bitrate, and cloud gaming one-way latency, respectively—substantially narrowing the gap between traditional trace-based simulation and real network conditions.

Abstract

Current and future applications demand ultra-low latency and consistent throughput, yet frequently traverse 5G cellular networks, so cope with volatile packet dynamics, as 5G base station schedulers dynamically react to user workloads and wireless channel conditions. The task of evaluating network algorithms in these environments is hamstrung by current tools: record-and-replay emulators sever the feedback interaction that exists between application end points and a commercial operator's proprietary 5G scheduler, while full-stack simulators rely on overly simplistic scheduling logic. To bridge this reality gap, we present NeuralEmu, a high-fidelity, machine learning-based emulation framework that learns complex 5G scheduler resource allocation behaviors directly from extremely high-resolution network telemetry tools. The first emulator to handle multiple clients, NeuralEmu utilizes machine learning to dynamically predict resource block allocations and modulation schemes based on instantaneous user buffer occupancy and channel states. To capture realistic cross-user contention, a traffic reconstruction model inverts cellular network scheduling results to recover the underlying traffic patterns of uncontrolled background users. Implemented as an high-performance Linux middlebox emulator, NeuralEmu reduces emulation error relative to the state of the art for various network applications including but not limited to 55% for web-page load time, 57% for WebRTC encoder bit rate, and 51% for cloud gaming packet one-way delay, providing an accurate, standardized testing ground for tomorrow's real-time interactive network protocols and applications.

Problem

• 5G network emulation
• scheduler fidelity
• feedback interaction
• resource allocation
• network dynamics

Innovation

• NeuralEmu
• 5G network emulation
• machine learning-based scheduling
• high-fidelity simulation
• traffic reconstruction