User Plane Congestion Control by User Class in 5G Sessions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wireless networks like 5G NR face challenges in managing user plane congestion without differentiating between user classes and data session types, leading to uniform throttling that affects all users and sessions equally, which can impact service quality and efficiency.
Innovation Solution
Implement a user plane congestion control service that classifies users into user classes based on service profiles and dynamic criteria, using AI/ML logic to adaptively manage congestion by applying differentiated throttling levels for uplink and downlink traffic, ensuring optimal resource utilization and quality of service.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If uniform congestion control is applied to all users, then network resource utilization is simplified to manage, but service quality differentiation between user classes deteriorates
Solution Approach 1:
The patent segments users into different user classes (e.g., premium users, ordinary users, protected users, non-protected users) and applies differentiated congestion control policies to each class. This segmentation allows the system to maintain simpler overall management while providing differentiated service quality, as each class has pre-defined thresholds and actions rather than requiring complex real-time individual assessments.
Solution Approach 2:
The patent implements local quality by applying different congestion control parameters (thresholds, actions, notification methods) to different user classes based on their specific service level agreements and priority levels. Each user class receives tailored congestion control treatment rather than a blanket uniform approach, enabling service quality differentiation while maintaining manageable complexity through standardized class definitions.
2Reliability
If differentiated congestion control is applied to user classes, then service quality is improved, but system complexity increases
Solution Approach 1:
The system reduces complexity by segmenting users into discrete classes with pre-configured parameters. Rather than managing individual user configurations, the system maintains a finite set of user class definitions with associated thresholds and actions, making the differentiated control approach manageable through standardized class templates.
Solution Approach 2:
The patent manages system complexity by changing parameters at the user class level rather than at the individual user level. Congestion control thresholds, actions, and notification methods are defined as class-level parameters that can be adjusted without modifying individual user configurations, thereby maintaining service quality differentiation while simplifying system management.
3Reliability
If congestion control thresholds are lowered to protect network resources, then network resource protection is improved, but user experience deteriorates
Solution Approach 1:
The patent applies local quality by setting different congestion control thresholds and actions for different user classes. Premium and protected users have higher thresholds and more favorable actions (e.g., notification before throttling), while non-protected users have lower thresholds and more aggressive control measures. This allows the system to protect network resources effectively while minimizing negative impact on high-value users.
Solution Approach 2:
The system implements preliminary anti-action by notifying certain user classes (e.g., protected users) before applying congestion control measures. This advance notification allows users to take corrective actions (e.g., reducing data usage, prioritizing critical traffic) before throttling begins, thereby protecting network resources while preserving user experience through proactive rather than reactive control.
4Productivity
If AI/ML logic is implemented for adaptive congestion management, then congestion control effectiveness is improved, but processing overhead increases
Solution Approach 1:
The patent reduces processing overhead by segmenting users into classes with pre-defined congestion control parameters. Rather than applying complex AI/ML algorithms to every individual user's traffic patterns, the system uses machine learning to optimize class-level thresholds and policies, then applies these pre-computed parameters to individual users through simple rule-based enforcement, significantly reducing real-time processing requirements.
Solution Approach 2:
The system implements preliminary action by using AI/ML logic to pre-compute optimal congestion control parameters for each user class based on historical data and network conditions. These pre-computed thresholds and policies are stored and applied during congestion events without requiring real-time AI/ML processing, thereby improving congestion control effectiveness while minimizing processing overhead during actual congestion management.
Data Source
AI summary
A method, a device, and a non-transitory storage medium are described in which a user plane congestion control service is provided. The service may analyze congestion information of user plane traffic in data sessions associated with a plurality of end devices. The service may include selecting, based on the analyzing, configuration parameters for uplink/downlink traffic mapped to one or more user class values of the end devices. The service may include generating a data session update indicating a modification to the configuration parameters for a user class value.