Dynamic Network Traffic Classification via Centralized Policy Server
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network traffic classification methods are inflexible and require manual configuration across multiple nodes, leading to inefficiencies and inaccuracies due to differing data models and information availability, resulting in high operational costs and suboptimal service application.
Innovation Solution
A dynamic classification and grouping technique that uses abstract classifications and stateful information to match traffic flows with predefined patterns, allowing for flexible service application across multiple nodes, with indications of membership transmitted between nodes to ensure consistent service treatment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual configuration of subscriber classification definitions is performed on each network node to ensure consistent traffic flow treatment, then service consistency across nodes is improved, but operational costs and configuration time increase
Solution Approach 1:
The patent introduces a centralized classification policy server as an intermediary that stores master classification definitions and provides them to network nodes on demand. This mediator enables consistent service treatment across nodes without requiring manual configuration at each node, thereby reducing operational time while maintaining reliability.
Solution Approach 2:
The system pre-configures classification definitions on a centralized policy server before they are needed at individual network nodes. When a node requires classification information, it retrieves pre-prepared definitions from the policy server, eliminating the need for time-consuming manual configuration at each node while ensuring service consistency.
2Reliability
If static pre-configuration data is used to define classification groups and allocate subscribers, then service treatment consistency is improved, but adaptability to service changes and subscriber modifications worsens
Solution Approach 1:
The patent transforms the static classification configuration into a dynamic system where classification definitions are stored centrally and can be modified in real-time. The policy server allows operators to update classification criteria and subscriber allocations dynamically, and these changes are automatically propagated to network nodes, maintaining service consistency while enabling rapid adaptation to changes.
Solution Approach 2:
The system enables dynamic modification of classification parameters (such as service thresholds, subscriber attributes, and grouping criteria) without requiring system reconfiguration. The policy server stores parameterized classification definitions that can be adjusted to reflect changing service requirements while maintaining consistent treatment across the network.
3Measurement precision
If equipment vendors develop custom software to match network operator classification schemas, then classification accuracy is improved, but device complexity and cost increase
Solution Approach 1:
Instead of requiring each network node to implement custom classification logic, the patent uses a centralized policy server that stores master classification definitions. Network nodes simply retrieve and apply these definitions as copied from the central repository, achieving accurate classification without the complexity of custom software development at each node.
Solution Approach 2:
The centralized policy server provides a universal classification service that can be used by multiple network nodes with different functions (SBCs, application servers, session controllers). This single universal system replaces the need for each node to have vendor-specific custom software, reducing overall system complexity while maintaining classification accuracy.
4Ease of operation
If information about traffic flow classification is distributed across multiple nodes with different data models, then node autonomy is improved, but classification consistency and service accuracy worsen
Solution Approach 1:
The patent merges the classification data storage and management function into a centralized policy server that consolidates classification definitions from multiple sources. Individual nodes retain their operational autonomy but retrieve unified classification information from the central server, ensuring consistent and accurate classification across the distributed network without sacrificing node independence.
Data Source
AI summary
Methods and apparatuses, including computer program products, are described for applying service based on classification and grouping of traffic flows. The method includes receiving a traffic flow, and matching the traffic flow to classification groups. The matching includes determining a first event associated with the traffic flow, comparing attributes of the first event with entry criteria of the classification groups, and assigning the first event to one or more classification groups where the first event meets the entry criteria of the one or more classification groups. The method includes identifying one or more service definitions for the traffic flow based on the classification groups assigned to the traffic flow, reconciling the one or more service definitions for the traffic flow, and providing a service to the traffic flow based on the reconciled service definitions.