Layer 3 IP Network Congestion Control via 5-Tuple Source Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network congestion control mechanisms are limited to layer 2 Ethernet networks, as they rely on flow identifiers and source MAC addresses, which are lost in layer 3 IP networks, preventing effective congestion notification and rate adjustment in these environments.
Innovation Solution
A method and device that detect network congestion by obtaining a 5-tuple of the congesting packet, generating a congestion control message carrying this information, and sending it to the source node or access forwarding device, allowing for rate reduction in layer 3 IP networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a congestion notification message (CNM) mechanism is used in a layer 2 Ethernet network with Flow-ID and source MAC address, then congestion control is effective, but the mechanism cannot be applied to layer 3 IP networks where packet headers are replaced at each routing
Solution Approach 1:
The patent changes the identification parameters from layer 2-specific fields (Flow-ID, source MAC address) to layer 3-applicable fields (5-tuple: source IP address, destination IP address, source port, destination port, protocol type). This parameter transformation enables the congestion control mechanism to function reliably across both layer 2 and layer 3 networks by using parameters that persist through routing operations.
2Adaptability or versatility
If packet headers are replaced at each routing in a layer 3 IP network, then routing flexibility is improved, but the source MAC address and Flow-ID are lost preventing congestion notification
Solution Approach 1:
The patent introduces an intermediary mapping mechanism that translates layer 3 packet information (5-tuple) into congestion control message format. This intermediary layer preserves source identification information through routing by mapping the 5-tuple from intercepted packets to generate appropriate congestion notifications, bridging the gap between routing flexibility and information preservation.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
Embodiments of the present invention provide a network congestion control method, a device, and a system. The method includes: receiving a first congestion control message, where the first congestion control message carries a 5-tuple of a packet that causes congestion; obtaining a flow identifier of the packet based on the 5-tuple of the packet that causes the congestion; generating a second congestion control message, where the second congestion control message carries the flow identifier corresponding to the 5-tuple; and sending the second congestion control message to a source node of the packet. In the embodiments of the present invention, a network congestion problem of a layer 3 IP network can be resolved. In technical solutions provided in the embodiments of the present invention, when congestion is detected in the network, the 5-tuple of the packet that causes the network congestion is obtained, the flow identifier of the packet that causes the network congestion is obtained based on information about the 5-tuple, and then a congestion control message is generated and is sent to the source node of the packet, to instruct the source node to reduce a rate of a packet that has the flow identifier. The technical solutions may be applied to a layer 3 network.