Streaming Quality Feedback Metrics for Bandwidth Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current quality feedback metrics in streaming services, such as those defined by 3GPP Tdoc S4-040073, provide only a coarse characterization of playback quality and fail to differentiate between frame types, leading to inaccurate representation of user experience, especially in cases where synchronization between audio and video is lost.
Innovation Solution
Introduce refined quality feedback metrics, including frame-rate-deviation, jitter-duration, synchronization-loss-duration, and number-of-corrupted-scene-cut-frames, which are reported with timestamps to provide a more detailed assessment of playback quality, allowing for differentiation between critical and less important frame types, and only report corruption duration if it exceeds a predefined value to optimize bandwidth usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coarse quality feedback metrics are used, then bandwidth usage is minimized, but measurement precision of playback quality is insufficient
Solution Approach 1:
The patent segments quality feedback into hierarchical levels: coarse metrics (initial buffering time, rebuffering duration, corruption duration) for bandwidth efficiency, and fine-grained metrics (frame-rate-deviation, jitter-duration, synchronization-loss-duration, number-of-corrupted-scene-cut-frames) for precision. This segmentation allows selective reporting based on quality thresholds, resolving the contradiction between measurement precision and bandwidth usage.
Solution Approach 2:
The patent introduces multiple quality parameters with different granularities. Coarse parameters provide overall quality assessment with minimal bandwidth, while fine parameters (frame rate deviation, jitter duration, sync loss duration, corrupted scene cut frames) provide detailed analysis only when quality degradation exceeds thresholds, optimizing the precision-bandwidth tradeoff.
2Measurement precision
If all quality metrics are reported continuously, then quality assessment accuracy is improved, but bandwidth consumption increases
Solution Approach 1:
The patent applies partial action by reporting only the necessary quality metrics based on current playback conditions. When quality degradation exceeds predefined thresholds, detailed metrics (frame-rate-deviation, jitter-duration, synchronization-loss-duration, number-of-corrupted-scene-cut-frames) are reported; otherwise, reporting is minimized or stopped, optimizing bandwidth usage while maintaining assessment accuracy when needed.
Solution Approach 2:
The patent implements quality-driven feedback mechanisms where quality metrics trigger conditional reporting. The system continuously monitors playback quality and activates detailed metric reporting only when degradation is detected, creating an efficient feedback loop that balances assessment accuracy with bandwidth conservation.
3Measurement precision
If frame type differentiation is implemented, then quality feedback accuracy is improved, but device complexity increases
Solution Approach 1:
The patent segments frame types into critical categories (scene cut frames vs. other frames) and applies different quality assessment rules to each. This segmentation enables differentiated quality feedback without requiring complex per-frame analysis, reducing device complexity while maintaining accuracy for visually significant frames.
Solution Approach 2:
The patent applies local quality assessment by treating different frame types with different quality thresholds and reporting priorities. Scene cut frames, which have higher visual impact, receive enhanced quality monitoring and reporting, while other frames use standard assessment, optimizing the balance between accuracy and complexity.
Data Source
AI summary
A method for quality feedback in a streaming service is shown, wherein at least one media stream is streamed to a client for playback, the method comprising determining a quality feedback value according to at least one quality metric from a pre-defined set of quality metrics, wherein the pre-defined set of quality metrics contains a quality metric related to at least one of the playback of the at least one media stream and a corruption of one out of at least two types of frames contained in the at least one media stream, and reporting the quality feedback value to a server. A system, a client, a server, a computer program, a computer program product and a protocol for quality feedback are also shown.