Server-Based Video QoE Calculation Reducing STB Processing Load
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for evaluating video Quality of Experience (QoE) in network services, particularly in video surveillance systems, burden set-top boxes (STBs) with high processing loads due to continuous monitoring and calculation of video stalling duration ratios, leading to performance degradation and increased costs.
Innovation Solution
A data processing method where network Key Performance Indicators (KPI) data is collected from network devices and used by a server to calculate video QoE, reducing the load on STBs and eliminating the need for embedded probes in all user-side STBs, by employing an associated model trained on historical data to determine video QoE based on network KPI and performance data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the probe continuously detects video stalling duration ratios and the STB calculates video QoE based on performance data and stalling duration ratios, then accurate video QoE can be obtained, but the processing load on the STB significantly increases
Solution Approach 1:
The patent extracts the QoE calculation function from the STB and relocates it to the server. The STB only needs to collect basic performance data and send it to the server, while the server performs the complex QoE calculation using the collected data along with historical information and algorithms. This extraction resolves the contradiction by maintaining measurement accuracy while significantly reducing the processing load on the STB.
Solution Approach 2:
The patent introduces a server as an intermediary between the STB and the QoE evaluation process. The server acts as a mediator that receives performance data from the STB, combines it with historical data and algorithms, and produces the final QoE evaluation. This intermediary approach allows the STB to maintain low processing load while still achieving accurate QoE measurement through the server's computational capabilities.
2Measurement precision
If probes are embedded in all user-side STBs to collect video stalling duration ratios, then comprehensive video QoE data can be gathered, but deployment scale and costs increase
Solution Approach 1:
The patent extracts the probe functionality from individual STBs and consolidates it into a centralized server-based solution. Instead of embedding probes in every STB, the server collects performance data from STBs and performs centralized analysis. This extraction eliminates the need for extensive probe deployment while maintaining comprehensive data collection capabilities through the server's centralized processing.
Solution Approach 2:
The patent creates a universal server-based platform that serves multiple STBs simultaneously. The server performs the function that would otherwise require individual probes in each STB, providing a multi-functional solution that reduces deployment complexity. This universal approach allows comprehensive QoE data collection across multiple devices without increasing per-device complexity.
Data Source
AI summary
A data processing method, a server, and a data collection device are provided. The method includes: obtaining, by a server, first network key performance indicator KPI data and performance data of a first set-top box STB, where the first network KPI data is network KPI data of a first video service stream, and the first STB is an STB that receives the first video service stream; and calculating, by the server, first video quality of experience QoE of the first video service stream based on an associated model, the first network KPI data, and the performance data of the first STB, where the associated model is a model obtained through training based on historical data, and the associated model is used by the server to calculate video QoE based on network KPI data and performance data of an STB.