Distributed UPF Re-Anchoring for Low-Latency Edge Service Selection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing edge computing systems face challenges in efficiently transitioning from centralized to distributed anchor points for user plane functions (UPFs) in mobile networks, particularly when Recursive Resolvers are not topologically close to the query originator, leading to suboptimal selection of edge application servers.

Innovation Solution

A method for on-demand UPF re-anchoring triggered by DNS requests or application-level triggers, involving network functions like SMF and UPF, which determine the need for re-anchoring based on SLA and policies, establish a local UPF session, and manage the DNS query process to ensure the UE reconnects through the new anchor point.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a centralized UPF anchor point is used, then network management and control is simplified, but latency increases and edge computing performance deteriorates

Engineering Contradiction:
Improvenetwork management complexityVSAvoidlatency
Core Design Contradiction:
Device complexityVSLoss of time

Solution Approach 1:

The patent segments the UPF functionality by introducing a distributed UPF instance at the edge network location, separating the anchor point function from the centralized control plane. This allows the SMF to remain centralized for management simplicity while the distributed UPF reduces latency by being topologically close to UEs

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds a spatial dimension to the UPF deployment by placing UPF instances at multiple network locations (edge, core, cloud) rather than a single centralized location. This multi-dimensional deployment enables latency optimization without sacrificing centralized control capabilities

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Loss of time

If a distributed UPF anchor point is deployed at edge locations, then latency is reduced and edge computing performance improves, but network complexity and UPF coordination overhead increases

Engineering Contradiction:
ImprovelatencyVSAvoidnetwork architecture complexity
Core Design Contradiction:
Loss of timeVSDevice complexity

Solution Approach 1:

The patent creates a universal UPF architecture where a single UPF instance can serve multiple functions and locations. The UPF can act as anchor point for local UEs, breakout point for edge services, and relay for cloud services, reducing the need for separate specialized components and simplifying overall network complexity

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The SMF acts as an intermediary that manages the complexity of coordinating multiple UPF instances. It handles session management, UPF selection, and traffic routing decisions, shielding edge locations from the complexity of inter-UPF coordination while enabling distributed deployment

Inventive Principle:
Principle #24Intermediary (Mediator)

3Ease of operation

If DNS-based server selection is used, then service discovery is simplified, but suboptimal server selection occurs when Recursive Resolvers are not topologically close to the query originator

Engineering Contradiction:
Improveservice discovery simplicityVSAvoidserver selection accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent performs preliminary UPF re-anchoring actions before DNS query resolution. The SMF proactively re-anchors the UPF to the edge location based on UE mobility and service requirements, ensuring that when the DNS query is later resolved, the topologically appropriate edge application server is selected with high accuracy

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces the mechanical DNS-based geographic selection mechanism with a network-controlled UPF re-anchoring mechanism. Instead of relying on DNS servers to guess the optimal location based on resolver IP addresses, the network proactively positions the UPF anchor point based on accurate UE location and service requirements

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12500860B2Application triggered setup of distributed anchor for edge computing
Publication Date: 2025.12.16 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)
  • US12500860B2 patent drawing
  • US12500860B2 patent drawing
  • US12500860B2 patent drawing

AI summary

Embodiments describing methods performed by a system comprising Network functions including at least a Session Management Function and a User Plane Function for performing a UPF re-anchoring for an application triggered at a User Equipment over an established Packet Data Unit (PDU) Session, where when the system receives the initial trigger for a service triggered by the application (e.g., a DNS request or an application level request), a determination is made by the system based on the information related to the application that re-anchoring of the UPF for that application into a local UPF closer to the UE is required. When re-anchoring of the UPF is successful, the initial trigger is discarded, and the UE is instructed to establish a new PDU session over the local UPF for that application.