Virtual Resource Automatic Selection via Supervised Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network virtualization techniques require manual and complex processes for virtual resource selection, leading to inefficiencies in resource allocation and quality of service assurance, especially during unexpected changes in network use environments such as emergencies or traffic fluctuations.
Innovation Solution
A virtual resource automatic selection system utilizing supervised learning to classify and allocate resources based on service quality requirements, traffic load, and failure status, enabling adaptive and proactive resource allocation across edge, core, and data center networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual and complex processes are used for virtual resource selection, then resource allocation can be performed with fixed policies, but processing time increases and labor burden increases
Solution Approach 1:
The system performs preliminary classification of virtual resource construction requests into multiple clusters based on service requirements, traffic load, and failure status. By pre-establishing classification boundaries through supervised learning with training data, the system prepares resource allocation decisions in advance, enabling rapid response when actual allocation is needed without manual intervention during emergencies.
Solution Approach 2:
The system automatically selects virtual resources through machine learning classification without requiring manual calculation or human intervention. The supervised learning model autonomously determines the appropriate virtual resource cluster based on input parameters, eliminating labor burden and reducing processing time while maintaining reliable quality of service assurance.
2Ease of manufacture
If fixed virtual resource selection policies are used, then implementation is simpler, but the policy may not be optimal when unexpected changes in network use environment occur
Solution Approach 1:
The system dynamically adapts to changing network environments by using supervised learning classification that can process varying input parameters including traffic load, failure status, and service requirements. The classification boundaries are established through training but can handle diverse and unexpected scenarios, providing both implementation simplicity through automated classification and adaptability to environmental changes.
Solution Approach 2:
The system changes its behavior based on input parameters by classifying requests according to multiple dimensions including service requirements, traffic load, and failure status. The supervised learning model adjusts resource allocation decisions based on the specific parameter values of each request, enabling optimal adaptation to different network conditions while maintaining a unified automated implementation approach.
3Reliability
If manual calculation processes are used to modify virtual resource selection policy, then quality of service can be assured, but labor burden increases and processing time increases
Solution Approach 1:
The system replaces manual calculation processes with an automated machine learning classification system. The supervised learning model performs the complex analysis and decision-making that would otherwise require manual intervention, eliminating labor burden while maintaining quality of service assurance through algorithmic resource selection based on multiple parameters including service requirements, traffic load, and failure status.
Solution Approach 2:
The system uses supervised learning with training data to establish classification boundaries that ensure quality of service. The model learns from historical data and automatically adjusts its classification decisions based on input parameters, providing continuous feedback-driven optimization without manual calculation. This automated feedback mechanism maintains reliable quality of service while eliminating the need for manual policy modification processes.
Data Source
AI summary
A virtual resource automatic selection system includes a setting unit that sets, for a plurality of pieces of training data composed of two or more parameters, a classification boundary between virtual resource clusters that are ranked in accordance with a capacity of a virtual resource in terms of a relationship between the two or more parameters. When a construction request for a new virtual network is received, a receiving unit receives information composed of the parameters. A determining unit determines to which of the virtual resource clusters the parameters belong. An allocating unit allocates a capacity of a virtual resource to the virtual network. An acquiring unit acquires network performance information from the virtual network. A second determining unit determines whether the network performance information satisfies a desired quality of service. The setting unit updates the classification boundary in accordance with a determination result of the second determining unit.