Intent-Driven Network Modeling for Automated Deployment Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network deployment methods are inefficient and error-prone due to reliance on manual analysis and operator experience, making it difficult to meet diverse and rapidly changing application scenarios.
Innovation Solution
A network management method and system that utilizes a knowledge graph-based logical network recommendation model to automatically deploy networks based on user intent, involving a logical and physical network model determination, simulation testing, and adjustment based on detection results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual service analysis and network requirement analysis are performed by operators, then network deployment can be completed, but the process is time-consuming, labor-consuming, and has high error probability
Solution Approach 1:
The patent replaces manual operator analysis with an automated system that uses machine learning models (logical network recommendation model and physical network recommendation model) to perform service analysis and network requirement analysis. This substitution of mechanical manual work with automated computational systems directly resolves the contradiction by dramatically improving deployment efficiency while reducing deployment time.
Solution Approach 2:
The system enables self-service automation where the network deployment system automatically performs service analysis, generates logical network models, determines physical network models, and executes deployment without requiring manual operator intervention at each step. This self-automating process resolves the productivity-time contradiction by making the system self-sufficient in performing deployment tasks.
2Reliability
If manual network deployment based on operator experience is used, then network can be deployed, but randomness is high and error probability is high
Solution Approach 1:
The patent changes the fundamental parameters of the deployment system by transitioning from experience-based manual parameters to data-driven automated parameters. The machine learning models use trained parameters and algorithms to systematically determine network configurations, eliminating the randomness and errors associated with manual operator experience while managing system complexity through structured model-based approaches.
Solution Approach 2:
By replacing manual operator judgment and experience with automated machine learning models, the system eliminates the variability and errors inherent in human decision-making. The logical network recommendation model and physical network recommendation model provide consistent, reproducible, and accurate deployment outcomes without the randomness of manual processes.
3Adaptability or versatility
If manual analysis is performed for each network requirement, then detailed network configuration can be achieved, but the process is inefficient and cannot meet frequently changing requirements
Solution Approach 1:
The patent introduces dynamic adaptability through machine learning models that can automatically adjust to changing network requirements. The system dynamically generates updated logical network models and physical network models based on new service requirements without requiring manual re-analysis, enabling the system to adapt to frequently changing requirements while maintaining high deployment efficiency.
Solution Approach 2:
The automated machine learning-based system replaces manual analysis processes, enabling the system to rapidly adapt to changing requirements through automated model regeneration and adjustment. This substitution allows the system to handle diverse and frequently changing network requirements efficiently without the bottlenecks of manual analysis.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
Embodiments of this application relate to a network management method. The method includes: obtaining a network type of a target network; obtaining a logical network model of the target network based on the network type and a logical network recommendation model; determining a physical network model of the target network based on the logical network model and a physical network recommendation model; and performing network configuration based on the physical network model of the target network. An advantage of the embodiments of this application lies in that, when a user inputs the network type of the target network instead of a large quantity of detailed network configurations, a network management system can automatically establish the required target network for the user, thereby greatly improving efficiency of establishing the target network.