"Realistic Dataset for V2X Evaluation and Prediction"

"The integration of accurate and realistic wireless network simulations that reflect real-world scenarios is crucial for advancing research in intelligent communications. By utilizing accessible simulation tools and datasets, one can avoid the expensive and time-consuming processes of gathering datasets. This can significantly lower the barrier for research with the purpose of validating novel machine learning algorithms for the network. We present a user-friendly dataset based on the interconnection of the Sionna ray-tracer with the ns-3 discrete network simulator with realistic vehicular-mobility. This dataset is particularly suited for testing algorithms in dynamic vehicular networks, urban deployments, and adaptive communication systems. Our work lowers the entry barrier for experiments with wireless network data, fostering reproducible research in machine learning-based quality of service prediction, network optimization, and intelligent communication system design."

能够精准还原真实应用场景的高保真无线网络仿真集成，对推动智能通信领域的研究具有关键意义。通过使用易于获取的仿真工具与数据集，可避免数据集采集环节的高成本与耗时问题，从而大幅降低验证面向无线网络的新型机器学习算法的研究门槛。本研究发布一款易用型数据集，其基于Sionna射线追踪器（Sionna ray-tracer）与ns-3离散网络模拟器（ns-3 discrete network simulator）的互联架构，并融入真实的车辆移动特性。该数据集尤其适用于动态车载网络、城市部署场景以及自适应通信系统中的算法测试。本研究降低了无线网络数据相关实验的入门门槛，助力基于机器学习的服务质量预测、网络优化及智能通信系统设计领域的可复现研究。