Evaluation of Uplink Video Streaming QoE in 4G and 5G Cellular Networks Using Real-World Measurements

Abstract

In this study, we propose a comprehensive method for evaluating the Quality of Experience (QoE) for uplink video streaming in commercial 4G and 5G cellular networks. Uplink video streaming is becoming increasingly important due to applications such as live content creation and remote operations. However, commercial networks are primarily optimized for downlink performance, posing significant challenges for achieving high uplink throughput. We investigate the factors affecting uplink QoE, including signal strength, network congestion, carrier aggregation, and coverage. Our methodology involves comprehensive data collection from network elements and streaming servers, encompassing RF parameters, IP packet data, and performance metrics such as throughput and latency. We employ full-reference metrics such as Video Multi-Method Assessment Fusion (VMAF), Structural Similarity Index (SSIM), and Peak Signal-to-Noise Ratio (PSNR) to assess video quality and address challenges in frame alignment for accurate evaluation. Our dataset includes video recordings, streamed video files, QoE metrics, IP packet data, and Radio Access Network (RAN) measurements, enabling robust regression analysis and machine learning for QoE prediction. We propose a system architecture for real-time data collection and streaming, integrating live video capture, streaming server setup, and network measurement tools. The results highlight the relationship between network conditions and video quality, demonstrating the impact of factors like path loss and Physical Resource Block (PRB) allocation and handover on QoE. This study provides valuable insights and practical solutions for optimizing uplink video streaming QoE in current and future cellular networks.