AI Uplink Routing Between Local and Remote UPFs in 5G
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The 5G mobile communication system experiences unsatisfactory delays due to the distance between the distributed unit (DU) and central unit (CU) sites, which limits the effectiveness of mobile edge computing and fails to meet millisecond-level latency requirements for real-time applications.
Innovation Solution
The system introduces an uplink route decision method using an artificial intelligence (AI) model to determine whether the user plane function (UPF) is located locally or remotely, allowing user data to be directly forwarded to local application servers via cell-site computing units, thereby reducing latency and transmission distance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If user data is transmitted through the traditional path (RU-DU-CU-Core Network), then network coverage and connectivity are ensured, but transmission latency increases and real-time application requirements cannot be met
Solution Approach 1:
An AI model is introduced as an intermediary component between the DU and UPF to automatically determine the optimal transmission path. The AI model receives information about UPF locations and UE requirements, then outputs routing decisions that balance latency reduction with network constraints, eliminating the need for complex manual routing configurations.
Solution Approach 2:
The system performs preliminary actions by pre-deploying UPFs at both local and remote locations, and pre-training AI models with network topology and traffic pattern data. This allows the system to make immediate routing decisions without real-time computation delays, as the AI model has already processed historical data and can quickly determine optimal paths.
2Loss of time
If UPF is deployed remotely at CU site, then network management is simplified, but transmission distance increases causing unsatisfactory delays
Solution Approach 1:
The UPF function is segmented into multiple deployment locations - both local UPFs at cell sites and remote UPFs at central locations. This segmentation allows the system to serve different traffic types differently: local UPFs handle latency-sensitive traffic while remote UPFs handle other traffic, achieving both low latency and simplified management.
Solution Approach 2:
The system dynamically selects between local and remote UPFs based on real-time conditions such as traffic type, network load, and UE requirements. The AI model continuously adapts routing decisions to current network states, allowing flexible deployment strategies that balance deployment simplicity with performance requirements.
3Loss of time
If UPF is deployed locally at cell site, then transmission latency is reduced, but network architecture complexity increases
Solution Approach 1:
The AI model acts as an intelligent intermediary that manages the complexity of having multiple UPF locations. It automatically handles the decision-making process for routing traffic to appropriate UPFs, shielding network operators from the operational complexity while enabling the performance benefits of local deployment.
Solution Approach 2:
The AI-based routing system provides universal functionality that can be applied across multiple cell sites and network configurations. Once deployed, the same AI model can manage routing for numerous local UPFs, reducing the operational complexity burden that would otherwise increase linearly with the number of deployed UPFs.
Data Source
AI summary
An uplink route decision method for a mobile communication system, wherein the mobile communication system determines an uplink path of a user equipment (UE) in the mobile communication system, includes obtaining an attribute of a user plane function (UPF) required by the UE, and determining whether the UPF is located in a local or a remote place according to an artificial intelligence model.