arXiv：cs.AI· Zifan Zhang, Mingzhe Han, Kannan Athreya, Yuchen Liu·· 5 小时前AI 评分43
Isaac-Net：面向大规模并行机器人学习的 GPU 批处理 5G 仿真
Network-in-the-Loop at Scale: GPU-Batched 5G Simulation for Massively Parallel Robot Learning
AI 导读
研究者提出 Isaac-Net，一个 GPU 批处理 5G NR 模块，可与 Isaac Lab 物理引擎同步推进数千个环境的上行链路仿真。其 NR 引擎复现了 ns-3 5G-LENA 的中位延迟，在闭环 32 机器人/环境下偏差 9%，并能还原信息年龄（AoI），而独立延迟模型会使 AoI 尾部轻约三倍。
正文
Abstract:Massively parallel GPU simulators train multi-robot policies in thousands of environments, and many fleets use private Fifth-Generation (5G) networks, where each robot's delay depends on its teammates' traffic. Network-in-the-loop training places a simulated 5G network inside this loop. However, GPU robot simulators reduce the network to an independent delay per message, while packet-level simulators run one scenario per CPU process and cannot keep pace with thousands of parallel environments. To bridge this gap, we present Isaac-Net, a GPU-batched 5G New Radio (NR) module that advances the uplink of thousands of environments in lockstep with Isaac Lab physics. Isaac-Net simulates every slot, the 0.5~ms interval in which the base station decides which robots transmit, for all environments at once. Extensive experiments confirm that its NR engine reproduces the median delay of ns-3 5G-LENA across loads, with a median delay 5--10\% low on an unseen carrier and 9\% high at 32 robots per environment in closed loop. The engine also reproduces the Age of Information (AoI), the age of each robot's newest delivered report, while an independent delay per message leaves the AoI tail about three times too light. In a configuration validated against 5G-LENA, Isaac-Net keeps the network in the loop for about one million robots on one GPU at 83\% of the Isaac Lab rate without the network, measured under a random policy. Isaac-Net is open source at this https URL
| Comments: | It is open source at this https URL |
| Subjects: | Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Robotics (cs.RO) |
| Cite as: | arXiv:2610.02370 [cs.NI] |
| (or arXiv:2610.02370v1 [cs.NI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02370 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zifan Zhang [view email]
[v1]
Thu, 1 Oct 2026 18:46:18 UTC (1,517 KB)
来源：arXiv：cs.AI · arxiv.org