Mesh Network Packet Classification via Port Identifier Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication technologies, particularly those adhering to the IEEE 802.11 Standard, face challenges in providing Quality of Service (QoS) in multi-hop distributed networks, failing to differentiate and prioritize various media qualities such as high-definition (HD) and standard definition (SD) videos effectively.
Innovation Solution
A method and apparatus for network-based traffic recognition and packet classification are introduced, which extract port identifiers from packet headers, utilize pre-defined and dynamic port tables, and application headers to classify packets, enabling intelligent packet management and prioritization within a multi-hop mesh network, thereby optimizing QoS by distinguishing different traffic and content categories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional QoS techniques are implemented on a single device in a single-hop network, then QoS management is simple and straightforward, but the system cannot handle multi-hop distributed networks and cannot discriminate different media qualities
Solution Approach 1:
The patent divides QoS management into multiple independent components distributed across different network nodes. Each node maintains local port tables and performs independent packet classification, rather than centralizing QoS management in a single device. This segmentation enables the system to handle multi-hop distributed networks while keeping individual node complexity manageable.
Solution Approach 2:
The patent introduces a new dimension of QoS management by implementing port-based classification at multiple hierarchical levels (individual nodes, mesh networks, and cloud services). This multi-dimensional approach allows the system to simultaneously support single-hop and multi-hop networks, as well as discriminate different media qualities, without proportionally increasing overall system complexity.
2Measurement precision
If packet classification is performed using detailed content analysis to distinguish HD and SD videos, then media quality discrimination is accurate, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary packet classification by examining port identifiers in transport layer headers before deeper content analysis. Port tables are pre-configured with media quality associations, allowing the system to quickly categorize packets as HD, SD, or other types based on port matching, avoiding the need to analyze actual video content for every packet.
Solution Approach 2:
The patent extracts and utilizes port identifier information from packet headers as a key feature for classification, separating this classification function from full content analysis. By taking out the port-based classification step, the system achieves accurate media quality discrimination without the computational overhead of analyzing video content itself.
3Measurement precision
If dynamic port tables are maintained and updated for each application, then traffic recognition accuracy is improved, but memory usage and table management overhead increase
Solution Approach 1:
The patent creates port tables that serve multiple functions: they are used for packet classification, media quality discrimination, and traffic prioritization. A single port table structure supports various applications and media types, reducing the need for separate data structures for each application while maintaining high recognition accuracy.
Solution Approach 2:
The patent implements port tables at local network nodes rather than maintaining centralized tables for all applications. Each node maintains only the port information relevant to its local traffic, reducing overall memory consumption while maintaining accurate local traffic recognition. The distributed nature of port tables allows each node to optimize its own memory usage based on local requirements.
Data Source
AI summary
An embodiment is a technique to perform network-based traffic recognition and packet classification. A port identifier in a transport layer header of a packet having a packet type associated with a priority level is extracted. The packet is transmitted from or to an application according to a network protocol in a multi-hop mesh network having a local node and a remote node. The port identifier includes a port number. The packet is classified into the packet type using the port identifier and one of a pre-defined port list, a dynamic port table, and an application header of the application.