Smart Buffering for OTT Video via User Behavior Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video buffering systems waste network resources as they do not account for individual user viewing behaviors, leading to inefficient use of radio resources and bandwidth, especially in over-the-top (OTT) video delivery.
Innovation Solution
Implementing a smart buffering system with machine learning that predicts video buffer sizes based on user viewing habits by analyzing traffic counters, such as video content type, watching time, pause, and rewind durations, to optimize buffer management and reduce unnecessary data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional video buffering systems use fixed buffer sizes, then video playback stability is improved, but network resource efficiency deteriorates due to unnecessary data transmission
Solution Approach 1:
The patent applies dynamics by transitioning from fixed buffer sizes to dynamic buffer size adjustment. The system continuously monitors user viewing behaviors (pause, rewind, fast-forward actions) and network conditions, then adapts buffer sizes in real-time. This allows the buffering system to optimize between playback stability and network efficiency by adjusting buffer capacity based on actual user needs rather than using a static predetermined size.
Solution Approach 2:
The patent implements parameter changes by modifying buffer size parameters based on observed user behavior patterns. The system tracks viewing habits and uses this data to dynamically adjust buffering parameters, changing the buffer capacity, refresh rates, and pre-buffering amounts according to individual user preferences and current network conditions, thereby resolving the contradiction between stability and efficiency.
2Reliability
If larger video buffers are used, then video quality of experience is improved, but bandwidth consumption increases leading to resource waste
Solution Approach 1:
The patent uses parameter changes by dynamically adjusting buffer size parameters based on user behavior analysis. Instead of consistently using large buffers, the system modifies buffering parameters (capacity, refresh rate, pre-fetch amount) according to observed user patterns, thereby maintaining video quality when needed while reducing bandwidth consumption during periods of lower user engagement.
Solution Approach 2:
The patent applies partial action by buffering only the necessary amount of video data based on user behavior predictions. Rather than always pre-buffering excessive data to ensure quality, the system buffers partial amounts tailored to individual user patterns, achieving adequate video quality experience while minimizing unnecessary bandwidth consumption.
3Device complexity
If uniform buffering strategies are applied to all users, then system complexity is reduced, but adaptability to individual user behaviors deteriorates
Solution Approach 1:
The patent implements self-service by enabling the buffering system to automatically learn and adapt to individual user behaviors without requiring complex manual configuration. The system monitors user actions (pauses, rewinds, fast-forwards), automatically analyzes patterns, and adjusts buffering strategies autonomously, thereby achieving high adaptability while keeping the user interface and configuration simple.
Solution Approach 2:
The patent uses feedback mechanisms by continuously monitoring user viewing behaviors and using this information to adjust buffering strategies. The system collects feedback on user actions, processes this data to identify patterns, and automatically modifies buffering parameters accordingly, creating a closed-loop system that adapts to individual users while maintaining manageable system complexity through automated feedback processing.
Data Source
AI summary
Aspects of the subject disclosure may include, for example, selecting a set of traffic counters, wherein the traffic counters provide a profile of viewing habits of a user, and wherein the traffic counters are extracted from video streaming by the user; predicting a size of a video buffer based on the traffic counters selected; and building the video buffer based on the predicted size. Other embodiments are disclosed.