Adaptive Bitrate Algorithm Cross-Period Quality Smoothing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video streaming systems, abrupt changes in video quality occur when transitioning between different profile ladders during a streaming session, degrading the playback experience.
Innovation Solution
The adaptive bitrate algorithm considers quality changes when selecting new profiles, smoothing transitions by analyzing changes in quality levels and profile levels, and adjusting reward values accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the adaptive bitrate algorithm selects profiles based solely on current network conditions and available bandwidth, then the responsiveness to network changes is improved, but abrupt quality changes occur when transitioning between profile ladders
Solution Approach 1:
The system performs preliminary analysis of quality changes before selecting a new profile. The quality change analysis module evaluates the difference between current and candidate profiles across multiple dimensions (resolution, bitrate, frame rate) and uses this analysis to adjust reward values, preventing abrupt quality transitions before they occur.
Solution Approach 2:
The system implements feedback by continuously monitoring the current profile's quality characteristics and using this information to influence future profile selections. The quality change analysis provides feedback on the magnitude of quality transitions, which adjusts the reward values to penalize large quality changes and maintain quality consistency.
2Adaptability or versatility
If the profile ladder changes during different periods of a streaming session, then the adaptability to different content requirements is improved, but noticeable quality drops occur at period transitions
Solution Approach 1:
The system applies local quality adjustment by modifying the reward values specifically for profiles that would cause large quality changes during period transitions. Instead of uniformly treating all profile changes, the quality change analysis identifies problematic transitions and applies localized adjustments to the reward function, allowing the system to adapt to different content requirements while minimizing noticeable quality drops at transition points.
3Manufacturing precision
If the client selects higher bitrate profiles when bandwidth is available, then the video quality is improved, but the bitrate changes become more volatile across segments
Solution Approach 1:
The system changes the parameters used for profile selection by introducing quality change metrics (resolution difference, bitrate difference, frame rate difference) as additional factors in the decision process. These parameter changes are incorporated into the reward value calculation, causing the system to balance quality improvement with bitrate stability when selecting profiles.
Data Source
AI summary
In some embodiments, a method calculates a reward function for an adaptive bitrate function that is based on a bandwidth and profiles in a profile ladder for a current segment of content. The reward function calculates a first profile should be selected for the current segment and, based on a previously selected profile from a prior segment and profiles in the profile ladder, changes the first profile to a second profile from the profile ladder. The method requests the second profile from the profile ladder for the current segment.