A Queueing Model for QoS and QoE Trade-off Analysis of XR Traffic in 5G/6G Networks

Extended reality applications combine haptic feedback and high resolution video, demanding both ultra-reliable low-latency communication and high data rates. Fifth-generation advanced and sixth-generation networks face the challenge of serving these traffic mixes under strict quality-of-service constraints while managing radio resources efficiently. Efficiently allocating limited wireless resources while maintaining both Quality of Service (QoS) and Quality of Experience (QoE) is therefore a key challenge. A discrete-time queueing model of a single base station serving one XR user addresses this gap. Haptic subframes arrive randomly and are served with strict priority. Video frames arrive periodically, each divided into subframes; when a new frame arrives, any unsent subframes from the previous frame are discarded. The channel quality evolves as a finitestate Markov chain with equal-probability SNR states. The system state tracks the number of served video subframes and the current channel condition. The periodic video arrival introduces a cyclic structure, leading to a time-inhomogeneous Markov chain over the frame period. Stationary probabilities are obtained by forward iteration until convergence. The haptic subframe success rate, video frame success rate, and average active resource blocks per transmission interval are derived from them. A video quality-of-experience metric based on delivered frame size is also computed.