遇见数据集

Application-Level Network Traces for Edge-Controlled Robotic Navigation – Wi-Fi/tshark Packet Captures (August 2025)

收藏
Zenodo2025-10-27 更新2026-05-26 收录
官方服务:

资源简介:

This dataset contains collected packet-level measurements from application-level communication monitoring conducted during Validation Scenario 2: Application Performance Evaluation, on 29 August 2025, at the WZL Testbed as part of the TARGET-X project, within the Manufacturing Vertical. The validation focused on assessing how network conditions affect real-time behavior of an edge-controlled mobile robot during navigation and planning phases of an assembly/disassembly task. The analysis was carried out using the 5G monitoring pipeline, extended with tshark to capture live communication packets exchanged between the robot and the edge server during task execution. The tests targeted two communication-critical segments of the disassembly workflow: Step 2 – Path planning to the assembly station Steps 7 and 10 – Navigation to and from the temporary station Each validation trial represented a full A–B–A navigation cycle, lasting approximately 45 seconds, during which ROS 2 topics (sensor data, localization, navigation feedback, and control commands) were transmitted over a wireless network (Wi-Fi) connecting the robot and the edge server. While the experiments were conducted over Wi-Fi, the setup and results are representative of 5G-based deployments, offering insight into the application-level network requirements for real-time robotic control. A total of 10 navigation trials were performed, complemented by a static baseline measurement with the robot idle. Each trial’s packet-level data were captured using tshark, enabling detailed analysis of bandwidth consumption, topic-level throughput, and latency behavior. Summary of Results (as reported): Mean uplink throughput: 36.56 Mbps (σ = 0.41 Mbps) Mean downlink throughput: 1.73 Mbps (σ = 0.01 Mbps) Stable low-variance behavior across trials, confirming sufficient bandwidth for real-time operation. Contents: Raw network capture files (.pcap) recorded via tshark Trial summaries and throughput statistics (.csv) Testbed: WZL, RWTH Aachen UniversitySoftware tools: tshark, ROS 2 Humble, Zenoh middleware, NAV2 stackNetwork setup: Wi-Fi (operational baseline for edge-controlled navigation, representative of 5G setup) This dataset supports ongoing TARGET-X research on application-level communication requirements for edge-controlled robotic systems, forming part of Work Package 2: Manufacturing Vertical Implementation.

提供机构:
Zenodo
创建时间:
2025-10-27
二维码
社区交流群
二维码
科研交流群
商业服务