Packet Delay Statistics Using Multi-Flag Sampling in Congested Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for high-precision delay statistical in packet transmission processes, particularly in 5G networks, face challenges in accuracy and reliability due to low sampling rates and packet loss during network congestion.
Innovation Solution
The implementation of an in-band flow information measurement (iFIT) technology using alternate-marking for passive and hybrid performance monitoring, which involves marking packets with delay measurement flags and recording timestamps at network elements, followed by centralized computation to calculate packet loss and delay indicators, and enhancing accuracy by distributing delay flags across multiple packets within a measurement period.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If alternate-marking method is used to mark packets for delay measurement, then measurement coverage is improved, but sampling rate decreases leading to lower measurement precision
Solution Approach 1:
The patent segments the measurement process by dividing packets into different types (first packets with delay measurement flags and second packets without flags). This segmentation allows the system to measure delay on specific packets while maintaining overall measurement coverage, resolving the contradiction between coverage and precision by selectively applying measurement to segmented portions of traffic.
Solution Approach 2:
The patent applies partial action by marking only certain packets (first packets) with delay measurement flags rather than all packets. This partial marking reduces the sampling burden and improves precision on marked packets while still providing adequate measurement coverage through strategic selection of which packets to mark, balancing coverage and precision.
2Difficulty of detecting and measuring
If packets are marked with delay measurement flags during network congestion, then delay detection capability is improved, but packet loss increases reducing reliability
Solution Approach 1:
The patent implements preliminary action by pre-marking packets with delay measurement flags before transmission during normal network conditions. This allows the measurement framework to be established in advance, so when congestion occurs, the marked packets are already prepared for measurement without adding processing burden during congestion, thereby maintaining reliability while improving delay detection capability.
Solution Approach 2:
The patent converts the harmful effect of network congestion into a benefit by using marked packets to specifically measure and identify delay issues during congestion events. The marked packets serve as probes that turn the congested network state into an opportunity for targeted delay measurement, transforming the harmful congestion condition into useful measurement data.
3Measurement precision
If multiple packets are marked with delay flags within a measurement period, then sampling rate increases improving measurement precision, but system complexity increases
Solution Approach 1:
The patent applies periodic action by organizing delay measurements into structured measurement periods with defined start and end times. Multiple packets are marked with delay flags at regular intervals within each period, creating a periodic measurement pattern. This periodic structure simplifies the system by providing a predictable rhythm to the measurement process, making it easier to manage complexity while achieving high precision through multiple samples per period.
Solution Approach 2:
The patent implements feedback mechanisms where measurement results from marked packets are collected and used to adjust or validate the measurement process. The system receives delay measurements from multiple marked packets, processes this feedback information, and uses it to maintain or improve measurement precision while managing system complexity through structured feedback loops that guide the measurement operation.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
This application discloses a delay statistical method, an apparatus, a storage medium, and a system. The method includes: obtaining a first packet and a second packet that belong to a same measurement period, where the first packet includes first flow detection information, the second packet includes second flow detection information; both the first flow detection information and the second flow detection information include a delay measurement flag, a first flag, and a second flag; the first flag of the first flow detection information indicates a first sequence number of the first packet in the measurement period; both the second flag of the first flow detection information and the second flag of the second flow detection information indicate a total quantity of packets having delay measurement flags in the measurement period; the first flag of the second flow detection information indicates a second sequence number of the second packet in the measurement period; sending, to a statistical apparatus, a first delay parameter obtained based on the first flow detection information; and sending a second delay parameter obtained based on the second flow detection information. This application can improve accuracy and reliability of delay statistical.