GTI发布XR测评方法白皮书-英-21页_1mb

GTI 5G Metrics and Test Methods Towards XR White Paper Summary

1. Introduction

• XR (Extended Reality, including VR, AR, MR) is a key application driving early metaverse development.
• 5G networks (large bandwidth, low latency, high reliability) are crucial for enabling high-quality XR experiences.
• Current subjective user test result comparisons for XR services are limited by diverse methodologies and indicators, necessitating standardized objective metrics and testing methods.

2. Service Quality Evaluation Metrics for XR

• Defined key Service Quality Indicators (SQIs) based on XR application characteristics and user experience (UX).
• Key Qualities:
  • Presentation Quality:
    • Stal Worry Period (total time)
    • Distortion Rate (time-weighted PDR calculation)
    • Black Border Rate (angle/intensity calculation for user/head movement)
    • Audio-Video Synchronization (time lag between component frames)
    • Audio-Motion/Speech Synchronization (time lag between audio and corresponding visual cues)
    • Instruction-Motion Synchronization (time lag between user input and character action)
  • Interaction Quality:
    • Operation Response Latency (time from user action to visible response)
    • Motion-to-Photons (MTP) Latency (end-to-end delay for head movement feedback)
    • Motion-to-High Resolution (MTHR) Latency (delay for resolution adjustments during head movement)
  • Source Quality:
    • Video Bit Rate
    • Video Frame Rate
    • Resolution
    • Field of View (FOV)

3. 5G Performance Metrics for XR

• Assessed the need for performance metrics tailored to XR’s frame-based transmission characteristics.
• Key Adjustments: Expanded the statistical unit from packet level to frame level for: Delay, Throughput, and Reliability.
• Metrics Introduction:
  • Frame Delay (RA N / Network Transmission / Handover / Jitter): Measured time intervals between e.g., frame reception/transmission by base station/equipment. Standard formula: Tend - Tstart.
  • Frame Throughput (RA N / Network Transmission): Calculated ratio of frame data size (M) to frame delay (Tend - Tstart). Formula: M / (Tend - Tstart).
  • Frame Reliability (RA / Downlink/ Uplink): Success rate of frame transmission (1 - E_lost / E_total).
• Objective and unified definitions and calculation methods are proposed for these XR-tailored metrics.

4. Equipment Capability Requirements & Test Methodology

• Terminal: Required capabilities include XR content playback/rendering, attribute recording (resolution, frame rate, bit rate), frame-level timestamp recording, HMD information recording, and specific performance metric tracking.
• Server: Must allow recording of frame-level timestamps (both Uplink and Downlink) and data packet sizes.
• Network Equipment (GNodeB/5GC/Stats Points): Requires support for frame-level timestamp and packet size recording to facilitate XR metrics calculations.
• Test Scenarios: Defined multiple real-world environmental scenarios.
• Test Services: Cov er various XR services like Cloud Gaming, VR Videos, AR Collaboration, AR Movies.
• Test Methodology: Standardized test processes, durations, device connection protocols, and data processing rules. Includes discarding outliers and application of statistical weighting during data aggregation.

5. Conclusions & Next Steps

• This report aims to define objective XR service quality evaluation metrics, establish 5G performance metrics suitable for XR, set equipment capabilities and test methods, enabling more consistent measurement and optimization of network performance related to the immersive XR user experience.