Distributed Network Selection Using Computing-Resource-Aware Node Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing distributed network selection methods for 6G networks fail to effectively and reasonably select network nodes due to the deep integration of computing-network and the coexistence of multiple network nodes.
Innovation Solution
A computing resources-based distributed network selection method where a terminal acquires or receives computing-resource information from a computing-resource orchestration center to determine a target RAN side and network node based on computing-resource information, enabling effective network node selection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a centralized network architecture is used, then network control is simplified and easier to manage, but it cannot uniformly meet diverse 6G service scenarios and network requirements
Solution Approach 1:
The patent divides the centralized network architecture into distributed network edge nodes (micro-cloud units) that can independently make local decisions. Each edge node segments the network control function, allowing diverse service scenarios to be handled locally while maintaining overall system coordination, thus resolving the contradiction between control simplicity and scenario adaptability.
Solution Approach 2:
The patent introduces a new dimension of distributed intelligence at the network edge, transforming the traditional single-dimensional centralized control into a multi-dimensional architecture where both centralized coordination and distributed autonomous decision-making coexist. This enables the network to adapt to diverse scenarios while maintaining manageable control through hierarchical organization.
2Ease of operation
If existing distributed network selection methods are used, then network node selection can be performed, but they cannot be applied under the circumstances of deep integration of computing-network and coexistence of multiple network nodes
Solution Approach 1:
The patent changes the selection parameters from traditional network-only metrics to computing-resource-based metrics. By introducing computing resource information (CPU, memory, storage, energy) as key selection parameters, the method becomes adaptable to computing-network integrated scenarios while maintaining network node selection functionality.
Solution Approach 2:
The patent creates a universal network selection method that works across multiple scenarios including traditional networks and computing-network integrated environments. The method universally applies computing-resource-based selection criteria regardless of the specific network architecture or service type, enabling broad applicability while maintaining effective node selection.
3Reliability
If traditional network selection methods are used, then network connectivity can be established, but effective and reasonable network node selection cannot be achieved in 6G distributed architecture with multiple network nodes
Solution Approach 1:
The patent implements feedback mechanisms where terminals report their computing resource requirements and receive recommendations for suitable network nodes. The system continuously monitors and adjusts node selections based on real-time computing resource states, improving both connectivity reliability and selection accuracy through iterative optimization.
Solution Approach 2:
The patent performs preliminary assessment of computing resource requirements before establishing network connectivity. By pre-evaluating terminal needs and matching them with appropriate network nodes based on computing capabilities, the system ensures both reliable connectivity and accurate node selection from the outset, rather than establishing connections first and optimizing later.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure provides a computing resources-based distributed network selection method and apparatus, and a readable storage medium. The method includes: acquiring computing-resource information of computing-resource nodes from a computing-resource orchestration center or receiving computing-resource information of computing-resource nodes sent by the computing-resource orchestration center; determining a target RAN side according to the computing-resource information of the computing-resource nodes, and sending a first message to the target RAN side, where the target RAN side is configured to determine a target network node to be accessed according to received computing-resource information of each network node in the first message or pre-configured computing-resource information of at least one network node locally, and send a second message to the target network node; and if a fourth message sent by the target RAN side is received, performing a network registration operation. The method enables reasonable and effective selection of network nodes, thereby effectively providing services for a terminal.