Application Feature Server Preprocessing for Rapid Third-Party App Deployment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for preconfiguring feature bases on Policy and Charging Enforcement Function (PCEF) and Traffic Detection Function (TDF) are slow to update, making it difficult to rapidly deploy third-party applications, especially when dealing with millions of services, due to storage limitations and the inflexibility in managing application feature fields.
Innovation Solution
An application feature server preprocesses application feature fields using hash algorithms and provides preprocessed fields along with an application ID to an Application Detection Control (ADC), enabling rapid deployment of millions of third-party applications by efficiently managing and updating application feature rules.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If preconfigured feature base is used on PCEF and TDF, then application detection capability is provided, but updating period is long and cannot meet rapid development requirement of third-party applications
Solution Approach 1:
The patent preprocesses application feature fields (extracting, hashing, and storing key characteristics) in advance in the RDB before applications are deployed. When a new third-party application needs to be deployed, the preprocessed feature base can be quickly matched against the application, eliminating the need for time-consuming real-time feature extraction and enabling rapid deployment.
Solution Approach 2:
The patent implements a dynamic feature base management system where the RDB can be updated with new application features as third-party applications are deployed. The system transitions from a static preconfigured feature base to a dynamic one that adapts to new applications, allowing the feature base to evolve with application development while maintaining rapid detection capabilities.
2Adaptability or versatility
If millions of application feature fields are stored to support millions of services, then detection coverage is improved, but storage space of PCEF and TDF becomes insufficient
Solution Approach 1:
The patent extracts only the essential and characteristic features from application feature fields (such as key URL patterns, IP addresses, port combinations) and stores these condensed representations in the RDB. This extraction process reduces the volume of stored data significantly while retaining the ability to accurately detect applications, thus solving the storage space limitation.
Solution Approach 2:
The patent creates a simplified copy of the application feature base in the RDB that contains only the essential detection rules and characteristics needed for application identification. This copied feature base is much smaller than the original comprehensive feature set but maintains sufficient detection capability for millions of services.
3Measurement precision
If comprehensive application feature fields are preconfigured, then detection accuracy is improved, but system complexity and storage requirements increase
Solution Approach 1:
The patent segments the application detection process into distinct components: feature extraction, feature hashing, rule generation, and matching. Each component handles a specific aspect of detection, which simplifies the overall system architecture and makes it easier to manage and maintain while preserving detection accuracy through the coordinated operation of these specialized modules.
Data Source
AI summary
A control method for application feature rules is provided. In the method, an application feature server preprocesses one or more application feature fields of a third-party application; and the application feature server provides the one or more preprocessed application feature fields and a corresponding application Identifier (ID) for an Application Detection Control (ADC) as an application feature rule.