Video Ladder Selection for Storage-Efficient Streaming Playback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of optimizing video storage while maintaining high-quality viewing services for diverse network conditions and device requirements, balancing storage costs with playback performance, is addressed by implementing an intelligent video lifecycle management system that uses machine learning to determine which video ladders to maintain based on user behavior and network conditions.
Innovation Solution
A system that integrates multiple strategies to monitor video file storage, evaluates video characteristics using trained models to estimate future viewing, and optimizes the storage of video ladders, reducing unnecessary storage while ensuring high-quality video delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple video ladders with different resolutions and bitrates are stored to support diverse network conditions and device requirements, then video quality and playback performance are improved, but storage costs and computational expenditures increase
Solution Approach 1:
The system dynamically adjusts the number of video ladders to maintain based on real-time network conditions and device capabilities. Instead of storing a fixed set of video versions, the system adapts the ladder configuration dynamically, retaining only the necessary number of versions (e.g., 2-3 ladders) that are most useful for current user conditions, thereby reducing storage costs while maintaining playback quality.
Solution Approach 2:
The system changes the parameters of video ladders by selectively removing or adding specific resolution and bitrate combinations based on predicted user needs. The machine learning model analyzes network conditions, device type, and viewing history to determine optimal ladder configurations, transforming the static storage approach into a parameter-optimized dynamic system that reduces overall storage requirements.
2Reliability
If a fixed number of video ladders are stored for each video to ensure adequate quality options, then playback performance is maintained, but storage resources are wasted on unnecessary video versions
Solution Approach 1:
The system implements feedback mechanisms where machine learning models continuously analyze user viewing behavior, network conditions, and playback performance data. This feedback is used to refine predictions about which video ladders are most likely to be requested, allowing the system to optimize storage configurations dynamically and reduce computational expenditures on analyzing and managing unnecessary video versions.
Solution Approach 2:
The system performs self-optimization by automatically determining which video ladders to retain based on predicted user needs without requiring manual intervention. The machine learning models self-adjust the ladder configurations based on accumulated data, enabling the system to autonomously reduce storage and computational resources allocated to video versions that are unlikely to be viewed.
3Quantity of substance
If video ladders are selectively removed to reduce storage, then storage costs decrease, but video stalls and rebuffering events may increase
Solution Approach 1:
The system performs preliminary actions by predicting future viewing patterns and network conditions before making storage decisions. The machine learning models analyze historical data and forecast which video ladders are most likely to be needed in the future, allowing the system to proactively retain only those essential versions and remove redundant ones, thereby reducing storage costs while preventing video stalls through advance planning.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for video storage optimization. One of the methods includes obtaining data for a plurality of video files of a content delivery system, wherein the plurality of video files corresponds to a set of video ladders, each video ladder identifying a respective transcoding version of video content represented by a video file of the plurality of video file and having different parameters; for each video file, executing one or more respective storage strategies to compute one or more respective output video scores, wherein each storage strategy uses trained models to evaluate characteristics of the respective video file and estimate future viewing of the video file; and in response to evaluating the output video scores computed for each of the plurality of video files, determining one or more actions to reduce storage for the plurality of video files.