ATELIER. QoE and Net.KPIs for On-Demand video transmission in Mobile Networks (emu.) part 2

This dataset contains all the raw (and processed) information for thousands of experiments that transfer a video on-demand from a core network to a UE emulated through SRSRan inside docker containers. The dataset contains both the network statistics but also the evaluation of the QoE through VMAF for each experiment. In total, 3 different videos has been used with different characteristics and different network impairments has been applied. Second part. Please download both parts of raw logs before unzipping. The logs contain network KPI information collected during the transfer of a video-stream using FFMpeg from the core net to a UE using SRS-RAN.Each experiment contains the full logs with the full network traces, for a total of more than 4000 experiments.The transferred and received videos are excluded due to the total weight. For a more detailed explanation on how the dataset has been generated, please refer to the associated article and GitHub repository.Title of the article: “ATELIER: Service Tailored and Limited-Trust Network Analytics Using Cooperative Learning”.本数据集包含数千个实验的原始（及处理后）数据，这些实验实现了视频点播从核心网络传输至通过Docker容器内的SRSRan（SRS无线接入网）仿真的用户设备 (User Equipment)。数据集既包含网络统计信息，同时也包含每个实验基于VMAF（视频多方法评估融合 (Video Multimethod Assessment Fusion，VMAF)）计算得到的QoE（体验质量 (Quality of Experience)）评估结果。本次实验共选用3个特性各异的视频，并施加了不同的网络损伤条件。 第二部分说明：请先下载原始日志的两个分卷后再进行解压。 该日志包含使用FFmpeg（FFmpeg）将视频流从核心网络传输至通过SRSRan仿真的用户设备过程中采集的网络KPI（关键绩效指标 (Key Performance Indicator)）信息。每个实验均包含完整日志与全量网络追踪数据，总计实验数量超过4000个。因文件总体量过大，已剔除传输与接收的视频文件。 若需了解该数据集的生成细节，请参阅相关学术论文与GitHub（GitHub）仓库。论文标题：《ATELIER：基于协同学习的服务定制化与有限信任网络分析》（ATELIER: Service Tailored and Limited-Trust Network Analytics Using Cooperative Learning）