Dynamic Traffic Classification Module Transitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic controllers are mass-produced with module collections that primarily identify frequent types of network traffic, leaving individual customers or small groups without the necessary modules to identify less-frequent traffic types, such as in-house software, which can be costly and require high technical expertise to customize.
Innovation Solution
The development of a Classification Module Language (CML) like CML_Tcl, which allows users to create customized traffic classification modules by using event-driven programming constructs, enabling users to write their own classification modules tailored to their specific network traffic needs, and a method for transitioning between different collections of traffic classification modules to accommodate changing traffic types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If traffic controllers are mass-produced with standardized module collections, then manufacturing cost is reduced and ease of manufacture is improved, but the ability to identify less-frequent traffic types deteriorates
Solution Approach 1:
The patent implements dynamic module collections that can be loaded and unloaded at runtime based on traffic type requirements. The system transitions from static, mass-produced module sets to dynamic, adaptable collections that can be customized for different customers' needs without requiring new hardware manufacturing.
Solution Approach 2:
The traffic controller is designed with a universal architecture that can handle multiple traffic identification scenarios through a single platform. By supporting interchangeable module collections, one device can serve diverse customers with different traffic type requirements, eliminating the need for custom-manufactured devices for each customer.
2Adaptability or versatility
If customized modules are created for each customer's specific traffic types, then adaptability is improved, but device complexity and technical expertise requirements increase
Solution Approach 1:
The system segments traffic identification functionality into independent, modular units. Each module handles a specific traffic type, and customers can select only the modules they need. This segmentation reduces overall system complexity by allowing selective activation of functionality rather than requiring all customers to have access to all possible traffic type identifiers.
Solution Approach 2:
The patent introduces an intermediary mechanism (the module collection management system) that mediates between the customer's traffic identification needs and the actual module implementation. This intermediary layer abstracts the complexity of module creation and management, allowing customers to benefit from customized identification without directly dealing with the underlying complexity.
3Adaptability or versatility
If module collections are updated to include new traffic types, then adaptability is improved, but the transition process and system complexity increase
Solution Approach 1:
The system employs dynamic loading and unloading of module collections, allowing updates to be applied at runtime without requiring system restart or complex transition procedures. The traffic controller can dynamically switch between different module collections based on current traffic identification needs, simplifying the update process.
Solution Approach 2:
The patent implements preliminary validation and preparation steps before module collection updates are applied. By checking compatibility and preparing transition procedures in advance, the system minimizes disruption during updates and reduces the complexity of the transition process itself.
Data Source
AI summary
Provided are computer-implemented methods and systems for transitioning between traffic classification modules. An example method for transitioning between traffic classification modules may include processing a plurality of packets associated with a plurality of sessions by a first collection of traffic-classification modules. The method may further include loading a second collection of traffic-classification modules. The method may continue with receiving one or more further packets flowing from a source network device to a destination network device. The one or more further packets may be associated with one or more new sessions. The method may further include processing the one or more further packets associated with the one or more new sessions by the second collection of traffic-classification modules. The method may continue with unloading the first collection of traffic-classification modules when no sessions of the plurality of sessions are associated with the first collection of traffic-classification modules.