Adaptive Video Transmission Control via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Real-time video streaming services face challenges in maintaining quality of service (QoS) due to unpredictable network bandwidth, as existing technologies lack effective methods to dynamically adjust bitrate and frames per second (FPS) based on varying network conditions.
Innovation Solution
A video transmission control method using machine learning to estimate the current and future network state, allowing for the selection of optimal video transmission parameters such as bitrate, FPS, and buffer flush rate, based on estimated network conditions, including bandwidth, stability, and communication provider information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If video transmission bitrate and FPS are increased to improve video quality, then video quality is improved, but network bandwidth consumption increases and may cause transmission failures in unstable networks
Solution Approach 1:
The patent implements dynamic adjustment of video transmission parameters (bitrate, FPS) based on real-time network state estimation. The system continuously monitors network conditions and adapts transmission quality accordingly, transitioning from static to dynamic parameter control to resolve the contradiction between video quality and transmission reliability.
Solution Approach 2:
The system changes transmission parameters (bitrate, frames per second) based on estimated network state. By adjusting these parameters dynamically according to network conditions, the system optimizes the balance between video quality and reliable transmission, preventing overload in unstable networks while maintaining high quality when network conditions permit.
2Measurement precision
If machine learning models use more feature data to improve network state estimation accuracy, then estimation accuracy is improved, but system complexity increases
Solution Approach 1:
The machine learning model processes network state estimation by segmenting features into different categories (transmission buffer features, network information features). This segmentation allows the system to handle complex multi-dimensional data in an organized manner, improving estimation accuracy while managing system complexity through structured feature processing.
Solution Approach 2:
The patent employs a universal machine learning framework that can process multiple types of features (buffer size, buffer duration, throughput, connection type, communication provider) through a single estimation model. This multi-functional approach enables accurate network state estimation across diverse network conditions without requiring separate specialized systems for each feature type.
Data Source
AI summary
A video transmission control method for adaptive data transmission includes predicting a network state through machine learning using information related to video transmission as a feature in a real-time video streaming environment; and determining an option for controlling video transmission on the basis of the predicted network state.