5G User Plane Delay Measurement via GTP Echo and Timestamps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 5G network performance measurements lack comprehensive monitoring of user plane data delay, particularly in terms of average and distributional metrics across different network slices and quality of service (QoS) classes, which hinders optimal resource allocation and user experience optimization.
Innovation Solution
Implementing advanced performance measurement techniques within the 5G network, including GTP echo and timestamp-based methods for QoS monitoring, to collect detailed delay data across various network slices and interfaces, enabling precise tracking of user plane data delays and optimizing network resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If comprehensive performance measurement techniques are implemented to monitor user plane data delay across different network slices and QoS classes, then measurement precision and information completeness are improved, but device complexity and implementation difficulty increase
Solution Approach 1:
The patent segments the performance measurement system into multiple specialized components: GTP echo function for basic delay measurement, timestamp-based measurement function for precise timing, per-network-slice measurement function for slice-specific monitoring, and per-QoS-class measurement function for quality-based categorization. Each component handles a specific aspect of measurement, reducing overall system complexity while improving precision.
Solution Approach 2:
The patent introduces intermediate measurement entities that act as mediators between the user plane data flow and the monitoring system. These intermediaries collect delay measurements at specific points in the data path (e.g., at the UPF or gNB) and forward processed measurement results to the monitoring entity, simplifying the architecture by centralizing complex measurement logic in dedicated intermediary components.
2Loss of information
If detailed delay measurements are collected across multiple network slices and interfaces, then information completeness is improved, but loss of time for measurement and processing increases
Solution Approach 1:
The patent implements preliminary action by embedding timestamps at the point of packet injection into the user plane and having network elements pre-configure measurement points and thresholds. This allows delay calculation to occur passively as packets flow through the network, rather than requiring active measurement interventions that would add processing time.
Solution Approach 2:
The measurement system utilizes self-service by having network elements automatically perform delay calculations using locally available timestamp information without requiring external measurement commands. Each network element (UPF, gNB) independently computes delay metrics based on embedded timestamps and forwards results autonomously, eliminating the need for centralized measurement control and reducing processing overhead.
3Measurement precision
If granular delay metrics are monitored for each network slice and QoS class, then measurement precision is improved, but the complexity of data processing and resource allocation increases
Solution Approach 1:
The patent applies local quality by configuring measurement parameters, thresholds, and reporting intervals specific to each network slice and QoS class according to their unique requirements. Critical services receive more frequent and precise measurements, while less demanding services use coarser measurement granularities, optimizing processing effort according to local quality requirements rather than applying uniform measurement complexity across all traffic.
Data Source
AI summary
Systems and methods for measurement and reporting of average UL and DL delays and delay distributions for an NG-RAN and UPF, as well as gNB-DU DL latency, for the N3 and N9 interfaces, internal UPF delay, and between the PSA UPF and NG-RAN and between the PSA UPF and UE are described. For the UL and DL average delay and delay distributions, the measurements are split into subcounters per 5QI/QCI, S-NSSAI and/or DSCPs, with the delay distributions subcounters in bins with discrete delay ranges. The measurements aon the N3 interface delay are also differentiated between a PSA UPF and an I-UPF.