Staged Traffic Classification for Broadband QoS
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data traffic classification methods in broadband communications networks face challenges in providing efficient, robust, and flexible QoS and bandwidth management, especially in shared bandwidth networks where dynamic and ephemeral port usage complicates early and accurate packet classification, leading to potential traffic congestion and delayed session handling.
Innovation Solution
A multi-stage traffic classification system that uses initial static classification methods at remote nodes, followed by dynamic adjustments based on deep packet inspection and signaling from aggregation nodes, allowing for flexible priority handling and accurate classification of data flows even at terminal nodes without requiring extensive processing power or specialized hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deep packet inspection is performed at terminal nodes for accurate traffic classification, then measurement precision of traffic flow is improved, but device complexity and processing power requirements increase significantly
Solution Approach 1:
The traffic classification process is segmented into multiple stages: initial classification at terminal nodes using simple port-based methods, followed by refinement at aggregation nodes using deep packet inspection. This segmentation allows accurate classification without requiring terminal nodes to have complex processing capabilities.
Solution Approach 2:
Aggregation nodes act as intermediaries between terminal nodes and the core network. They perform deep packet inspection on traffic flows and send signaling information back to terminal nodes, enabling accurate classification without burdening terminal nodes with complex processing requirements.
2Ease of operation
If static traffic classification methods are used at terminal nodes, then ease of operation is improved, but measurement precision of traffic flow deteriorates due to dynamic port usage
Solution Approach 1:
The system transitions from static port-based classification to dynamic classification through multi-stage processing. Initial static classification is performed at terminal nodes for simplicity, then dynamic refinement is applied at aggregation nodes using deep packet inspection and signaling mechanisms that adapt to changing traffic patterns and port usage.
Solution Approach 2:
Terminal nodes perform preliminary static classification using simple port-based methods before traffic reaches aggregation nodes. This preliminary action provides initial classification with minimal processing, while more accurate dynamic classification is performed later at aggregation nodes.
3Measurement precision
If multi-stage traffic classification with deep packet inspection signaling is implemented, then measurement precision is improved, but loss of time increases due to additional processing stages
Solution Approach 1:
Static classification is performed preliminarily at terminal nodes before traffic reaches aggregation nodes. This preliminary classification provides immediate initial routing decisions, while deep packet inspection signaling refines classification in the background, minimizing the time impact on traffic flow.
Solution Approach 2:
The system uses feedback signaling from aggregation nodes to terminal nodes to refine traffic classification. The signaling mechanism provides correction and refinement information without requiring re-processing of entire traffic flows, thus minimizing time loss while improving precision.
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
A system architecture and methods for data traffic flow classification are provided. An initial traffic class is assigned to a data flow as a current traffic classification, where the initial traffic class is based static traffic classification method(s) applied with respect to an initial packet of the data flow. A predetermined number of further packets of the data flow, subsequent to the initial packet, are analyzed based on predetermined factor(s), and a traffic class based on the analysis of the further packets is determined. The traffic class based on the analysis of the further packets is assigned as the current traffic classification of the data flow. Data indicating a traffic class for the data flow (based on a dynamic traffic classification method) is received, and the traffic class based on the dynamic traffic classification method is assigned as the current traffic classification of the data flow.