Analysis of Cellular Network Latency for Edge-Based Remote Rendering Streaming Applications

Recent advances in low-latency streaming cloud technology have enabled remote rendering applications such as Google Stadia and Occulus VR. Edge-clouds located close to the end-user are expected to play an important part in improving user experience for remote rendering application. In this paper, we investigate network latency in edge-based remote rendering over LTE networks. We quantify network latency and identify the roadblocks in deploying remote rendering on existing cellular networks with extensive measurements. We show that crossing traffic could severely deteriorate latency performance and network side resource management can solve the issue. We also find the impact of various radio configurations and scheduling algorithm in LTE networks. Our results shed light on how the network side and client side can cooperatively reduce network latency for emerging applications.