Quality of Experience Oriented Cross-layer Optimization for Real-time XR Video Transmission

TL;DR

This work tackles real-time XR video transmission over wireless networks by coupling MAC-layer scheduling with application-layer bitrate adaptation across two time scales. It introduces a cross-layer framework where the base station uses a frame-priority MS-DQN scheduler and the XR server employs a Transformer-based PPO (TPPO) bitrate adaptor, explicitly addressing the mismatch between per-slot resource allocation and bitrate adjustment periods. The approach yields QoE gains of up to 3.6%–37.8% overall and up to 49.9%–80.2% in transmission quality, validated through multi-user simulations and ablation studies against baselines. By leveraging frame-level XR semantics and historical context, the method provides practical improvements for multi-user XR QoE in dynamic wireless environments, while noting that the initial frame arrival distribution can significantly influence performance.

Abstract

Extended reality (XR) is one of the most important applications of beyond 5G and 6G networks. Real-time XR video transmission presents challenges in terms of data rate and delay. In particular, the frame-by-frame transmission mode of XR video makes real-time XR video very sensitive to dynamic network environments. To improve the users' quality of experience (QoE), we design a cross-layer transmission framework for real-time XR video. The proposed framework allows the simple information exchange between the base station (BS) and the XR server, which assists in adaptive bitrate and wireless resource scheduling. We utilize the cross-layer information to formulate the problem of maximizing user QoE by finding the optimal scheduling and bitrate adjustment strategies. To address the issue of mismatched time scales between two strategies, we decouple the original problem and solve them individually using a multi-agent-based approach. Specifically, we propose the multi-step Deep Q-network (MS-DQN) algorithm to obtain a frame-priority-based wireless resource scheduling strategy and then propose the Transformer-based Proximal Policy Optimization (TPPO) algorithm for video bitrate adaptation. The experimental results show that the TPPO+MS-DQN algorithm proposed in this study can improve the QoE by 3.6% to 37.8%. More specifically, the proposed MS-DQN algorithm enhances the transmission quality by 49.9%-80.2%.

Figures (13)

• Figure 1: XR transmission system framework. Real-time XR video is rendered on edge XR servers and then transmitted to clients for playback. BS uses frame information to schedule packets and improve transmission quality.
• Figure 2: Algorithmic diagram of the real-time XR transmission system.
• Figure 3: This example demonstrates three types of state transitions. In slot $t$, the agent of MS-DQN makes two scheduling decisions and transmits two frames.
• Figure 4: 2D-CNN-based neural network design. We estimate the value function and the advantage function respectively, then obtain the action value function.
• Figure 5: Convergence performance of MS-DQN and DDPG algorithms for wireless resource allocation problem.