Multimedia Encoding Rate Adjustment via QoE Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multimedia encoding rate adjustment methods in IP multimedia services rely solely on user feedback, leading to suboptimal Quality of Experience (QoE) due to inefficient encoding rate adjustments, which may result in lost information and inadequate network performance.
Innovation Solution
A method and system that adjust the multimedia encoding rate based on current network performance and user expectations, using a QoE evaluation model and multimedia resource decision model to determine an optimal encoding rate that balances media quality and network performance, thereby improving user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the encoding rate is increased or reduced only according to the QoE fed back by the user, then the user's QoE is improved, but the network performance may not meet the requirement and information is lost
Solution Approach 1:
The system pre-establishes a mapping relationship between QoE values and encoding rates through offline training. When a QoE value is obtained, the corresponding encoding rate is directly retrieved from the pre-built model, eliminating the need for multiple iterative adjustments and preventing information loss during the adjustment process.
Solution Approach 2:
A multimedia resource decision model acts as an intermediary between QoE feedback and encoding rate adjustment. This model, trained offline with network performance data, translates QoE values into appropriate encoding rates while considering network conditions, thus preventing information loss and ensuring network performance requirements are met.
2Reliability
If the encoding rate is adjusted multiple times to find a proper value, then the QoE value meets user expectation, but the adjustment process is complex and time-consuming
Solution Approach 1:
The system pre-trains a multimedia resource decision model offline that captures the complex relationship between QoE, network performance, and encoding rates. During online operation, the system directly queries this pre-trained model to obtain the optimal encoding rate in one step, eliminating multiple iterative adjustments and significantly reducing adjustment time.
Solution Approach 2:
The problem is divided into two independent phases: offline model training phase where complex computations are performed, and online query phase where simple lookups are executed. This segmentation moves the computationally intensive work to offline, making the online adjustment process fast and efficient.
3Productivity
If the encoding rate is reduced to relieve congestion, then network performance is improved, but video quality declines and user QoE deteriorates
Solution Approach 1:
The multimedia resource decision model serves as an intelligent intermediary that considers both network performance and video quality when determining encoding rates. It is trained offline with data showing the optimal balance between these conflicting factors, enabling it to select encoding rates that maintain video quality while adapting to network conditions.
Solution Approach 2:
The system dynamically changes the encoding rate parameter based on the optimal mapping learned during offline training. This mapping captures the precise relationship between network conditions, video quality requirements, and encoding rates, allowing the system to adjust the encoding rate parameter to achieve the best trade-off between network throughput and video quality.
Data Source
Figure 1~3
Figure 4
Figure 5
AI summary
A method, an apparatus, and a system for adjusting a multimedia encoding rate are disclosed. The method includes: obtaining an expected multimedia encoding rate by using the network available bandwidth corresponding to the transmission path of a multimedia service, the expected QoE of the multimedia service, and the current multimedia encoding rate as input parameters and according to a preset multimedia resource policy decision, where the expected multimedia encoding rate is used as a reference for adjusting the current multimedia encoding rate. With the present invention, under the current network performance, an optimal point for balancing the effects of the media encoding rate and the network performance on the QoE is found by adjusting the multimedia encoding rate, thus achieving optimal QoE. In this way, the adjustment process is simple and fast and the success rate is high, thus improving the QoE.