Video Quality Determination via RTT Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing OTT video quality evaluation methods are ineffective for TCP transmission due to the packet loss retransmission mechanism, which drastically reduces sending rates and results in low MOS-V values across the network, making it difficult to accurately determine video quality and locate network faults.
Innovation Solution
The method determines video quality using network layer KPI parameters, specifically by measuring RTT and packet loss rate with higher accuracy, and uses a mapping table to correct RTT for precise packet loss rate calculation, thereby isolating the impact of TCP packet loss retransmission on throughput determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If TCP packet loss retransmission mechanism is used, then packet loss can be recovered, but sending rate drastically drops and MOS-V values become extremely low across the network
Solution Approach 1:
The patent segments the network path into multiple measurement points and separates the measurement of packet loss rate from throughput determination. By measuring packet loss rate at specific points and using it to correct RTT values, the system isolates the impact of retransmissions from the overall quality assessment, allowing reliable packet loss detection without letting retransmission penalties drag down the entire network's evaluated productivity.
Solution Approach 2:
The patent introduces RTT correction as an intermediary mechanism. Instead of directly using raw RTT measurements that are contaminated by retransmission delays, the system corrects RTT values by compensating for packet loss rate effects. This intermediary correction layer separates the true network latency from retransmission-induced delays, enabling accurate quality assessment independent of TCP's reactive rate reduction.
2Measurement precision
If MOS-V values are used to evaluate video quality on all network devices, then comprehensive quality assessment is achieved, but fault location becomes impossible due to uniformly low values across the network
Solution Approach 1:
The patent applies local quality assessment by measuring packet loss rate and RTT at specific network devices along the transmission path rather than treating the network as a uniform whole. Each device's local packet loss rate is used to correct its specific RTT measurement, creating localized quality metrics that reflect actual conditions at each point. This enables identification of which specific network segment contributes to quality degradation.
Solution Approach 2:
The patent adds a new dimension to quality assessment by introducing packet loss rate as a separate measurement parameter independent of throughput-based MOS-V. Instead of relying solely on the single dimension of overall video quality (MOS-V), the system measures packet loss rate and RTT correction separately, creating a multi-dimensional quality space where fault location becomes possible through spatial variation analysis of these parameters across network devices.
3Reliability
If TCP sending rate is reduced due to packet loss, then retransmission can occur, but terminal device cannot receive sufficient packets for video playing
Solution Approach 1:
The patent performs preliminary measurement of packet loss rate before it fully impacts throughput determination. By measuring packet loss rate at intermediate points and using it to correct RTT values in advance, the system anticipates the impact of retransmissions and adjusts quality assessment accordingly, rather than reacting after packet quantity has already been reduced.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The present invention provides a method and an apparatus for determining video quality, and a method and an apparatus for locating a network fault. The method for determining video quality includes: obtaining a network key performance indicator KPI parameter on a first network device of a plurality of network devices, where the network KPI parameter includes a first round trip time RTT between a head-end device and the first network device; determining a Transmission Control Protocol TCP throughput of the first network device based on the network KPI parameter on the first network device; and determining video quality on the first network device based on the TCP throughput and a played video amount of the first network device. The present invention can accurately determine quality of a video transmitted by using the TCP protocol and accurately locate the network fault.