Network Optimization System Using ML Clustering for Streaming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing media streaming technologies face challenges in determining optimal network settings to improve quality of experience, as adjusting parameters like bitrate, buffer, and latency can be complex due to varying devices and networks, and require dynamic adjustments based on changing environmental conditions.
Innovation Solution
A network optimization system that uses clustering and classification algorithms to analyze network data from multiple devices, determine network groups, and send tailored treatment parameters to improve streaming quality, including adjusting settings such as bitrate and latency, without user intervention, and adapts to real-time changes in network conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If network and device settings are adjusted to improve streaming quality, then quality of experience is improved, but it becomes difficult to determine which parameters to adjust and how to dynamically change them
Solution Approach 1:
The system employs machine learning models that automatically analyze network conditions and device characteristics to determine optimal streaming parameters without user intervention. The models self-adjust settings based on real-time data, eliminating the complexity of manual parameter tuning while maintaining high streaming quality.
Solution Approach 2:
The patent dynamically changes streaming parameters (bitrate, resolution, buffer size) based on analyzed network conditions and device capabilities. The system continuously adjusts these parameters in response to changing environmental conditions, transforming static settings into adaptive, optimized values that resolve the contradiction between quality and complexity.
2Reliability
If manual adjustment of streaming parameters is attempted, then streaming quality may improve, but the vast number of devices and variety of networks make it difficult to determine optimal settings
Solution Approach 1:
The system segments the vast diversity of devices and networks into manageable clusters using unsupervised machine learning. By grouping similar devices and network conditions together, the system can determine optimal parameters for each segment rather than attempting to manually configure every possible combination, thus handling device and network variety effectively.
Solution Approach 2:
The patent introduces machine learning models as intermediaries between the diverse device/network environment and the streaming process. These models analyze device characteristics and network conditions, translating the complexity of variety into simplified, optimized parameter recommendations that work across different devices and networks.
3Reliability
If buffering is increased to prevent interrupted viewing, then quality of experience improves, but buffering time increases causing delays
Solution Approach 1:
The system dynamically adjusts buffer size based on real-time network conditions and device performance characteristics. Rather than using a fixed buffer size, the machine learning models continuously optimize buffer parameters to maintain viewing continuity while minimizing buffering time, adapting to changing conditions during playback.
Solution Approach 2:
The patent implements feedback loops where the system monitors actual streaming performance and uses this information to adjust buffer parameters. The machine learning models learn from observed buffering patterns and network behavior, continuously refining buffer size recommendations to achieve the optimal balance between viewing continuity and time efficiency.
Data Source
AI summary
Systems, methods, and computer-readable media are disclosed for optimizing network performance on a computer device to improve quality of experience by determining which network settings on the computing device to adjust. A clustering algorithm may identify various classes of networks and a classification algorithm may determine a network class specific to a network on a computing device. The effects of certain network settings for that networks class may be determined and the network setting and/or settings that optimizes the network performance may be promoted. The system may periodically analyze network data to recalculate the appropriate networks class and may determine different network settings based on the recalculation, facilitating mid-session improvements to the quality of experience.