Network Intent Decomposition with Measurement-Based Replanning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing intent decomposition systems face inefficiencies and low success rates due to the need to collect network states from a large number of intent operation objects with diverse vendors, leading to bottlenecks and difficulties in adapting to vendor changes.
Innovation Solution
A method where network devices exchange measurement values and re-decompose intents based on these values, shielding high-level systems from vendor-specific implementation details, thereby improving efficiency and success rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the intent system collects network states from a large quantity of intent operation objects with diverse vendors, then the intent achievement coverage is improved, but the system complexity and processing burden increase, causing bottlenecks
Solution Approach 1:
The patent introduces a standardized interface layer as an intermediary between the intent system and diverse vendor-specific intent operation objects. This interface abstracts vendor-specific implementations, allowing the intent system to interact with multiple vendors through a unified protocol without directly managing vendor-specific complexities, thus improving adaptability while controlling system complexity
Solution Approach 2:
The patent segments the intent operation objects into standardized functional modules that can be independently managed. By dividing the complex vendor-specific implementations into smaller, standardized units, the system can handle multiple vendors more efficiently, reducing the processing burden on the central intent system
2Loss of information
If the intent system directly interacts with vendor-specific implementations to obtain network states, then the data collection capability is improved, but the adaptability to vendor changes deteriorates
Solution Approach 1:
The standardized interface acts as a mediator that maintains data collection capability while insulating the intent system from vendor-specific changes. The interface translates vendor-specific data formats into a unified structure, ensuring continuous data collection regardless of vendor implementation changes
Solution Approach 2:
The patent creates a universal interface that can interact with multiple vendor implementations through a common protocol. This universal interface maintains the ability to collect data from diverse sources while adapting to different vendor changes through a single standardized entry point
3Measurement precision
If the intent system processes a large amount of network state data from multiple vendors, then the intent achievement accuracy is improved, but the processing efficiency deteriorates
Solution Approach 1:
The standardized interface serves as a filtering intermediary that pre-processes and standardizes vendor-specific data before it reaches the intent system. This reduces the data volume and complexity that the intent system must process, improving efficiency while maintaining accuracy through structured data transmission
Solution Approach 2:
The patent extracts and separates vendor-specific data processing logic from the core intent system. By extracting this functionality into independent interface modules, the system can process large amounts of data more efficiently while the core intent system focuses on high-level decision-making
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
This application provides an intent decomposition method and an apparatus. The method includes: A first network device sends a first sub-intent target value; the first network device receives a first message, where the first message is used to indicate that a first sub-intent target value is not achieved, and the first message includes a first measurement value of a second network device; and the first network device re-decomposes an intent based on the first measurement value, and sends first sub-intent target values obtained through re-decomposition, until all the first sub-intent target values are achieved or no first sub-intent target value is achieved. For a second network device that does not achieve the first sub-intent target value, re-decomposing the first sub-intent target value with reference to the first measurement value of the second network device can reduce a quantity of times of re-decomposing the first sub-intent target value, and improve efficiency. In addition, there is no need to pay attention to whether it is a multi-vendor scenario, so that a success rate and efficiency of intent achievement can be improved.