CGReplay: Capture and Replay of Cloud Gaming Traffic for QoE/QoS Assessment

TL;DR

Cloud gaming research is hampered by closed platforms and non-deterministic engines, which make reproducible QoE/QoS evaluation difficult. CGReplay captures uplink commands and downlink video frames in an ordered, synchronized action/reaction loop and replays them under controlled network conditions using UDP/RTP/H.264 with optional SCReAM. The open-source platform provides YAML-based configuration for server and player, and implements reliability mechanisms for command loss and frame loss (ACK/NACK with a sliding window $w$) while maintaining synchronization via $F_{ID}$ and $C_{ID}$. This work enables rigorous QoE assessment and provides a foundation for evaluating adaptive rendering and encoding strategies in real-world-like cloud gaming.

Abstract

Cloud Gaming (CG) research faces challenges due to the unpredictability of game engines and restricted access to commercial platforms and their logs. This creates major obstacles to conducting fair experimentation and evaluation. CGReplay captures and replays player commands and the corresponding video frames in an ordered and synchronized action-reaction loop, ensuring reproducibility. It enables Quality of Experience/Service (QoE/QoS) assessment under varying network conditions and serves as a foundation for broader CG research. The code is publicly available for further development.

Figures (6)

• Figure 1: Capturing online CG. The AP is connected to the Internet via a wired connection and connects to the laptop using Wi-Fi 6.
• Figure 2: Replaying CG traffic with CGReplay agents at server and player sides.
• Figure 3: Xbox Joystick Button Mapping with Corresponding Commands. The axes' continuous values range [-32767, 32767], and the button flag is 0 or 1.
• Figure 4: Action & Reaction Interactivity in CGReplay.
• Figure 5: Command loss scenario. {$C_1, C_2$} are lost through the network. Frame $f_3$ is retransmitted and the CG server waits for {$C_1$, $C_2$}.