Cross-Layer Congestion Marking via Data Link Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current congestion control mechanisms in data networks, such as ECN and PCN, operate independently in different layers of the protocol stack and do not effectively utilize early congestion detection from lower layers to prevent network congestion.
Innovation Solution
The method involves feeding congestion information from the data link layer, generated through mechanisms like BCN, upstream to the network layer for congestion marking, allowing for earlier detection and remedial action by using this information for higher-layer congestion control functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If congestion marking operates independently in different layers (ECN at network layer, BCN at data link layer), then each layer can perform its own congestion control functions, but the system cannot effectively utilize early congestion detection from lower layers to prevent network congestion
Solution Approach 1:
The patent implements feedback by having the data link layer send congestion information upstream to the network layer. The BCN mechanism at the data link layer detects congestion conditions and transmits this information to higher layers, enabling them to take preventive action. This cross-layer feedback loop resolves the contradiction by allowing early congestion detection while maintaining a manageable integration approach through standardized information exchange protocols.
Solution Approach 2:
The patent introduces a new dimension of interaction between protocol layers by enabling information flow from the data link layer (layer 2) to the network layer (layer 3). This vertical integration across dimensions allows congestion detection at the lower layer to influence higher layer decisions, preventing congestion before it occurs rather than reacting after independence was maintained.
2Loss of time
If early congestion detection is implemented at the data link layer, then preventive action can be taken before network congestion occurs, but the congestion information must be fed upstream to higher layers which increases protocol complexity
Solution Approach 1:
The patent applies preliminary action by detecting congestion at the data link layer before it propagates to the network layer and causing broader network issues. The BCN mechanism identifies early signs of congestion and triggers preventive marking actions upstream, reducing the time loss by acting before the problem escalates rather than waiting for network layer congestion to manifest.
Solution Approach 2:
The patent uses congestion information as an intermediary that bridges the data link layer and network layer. This information acts as a mediator, carrying congestion state from the lower layer to the higher layer without requiring direct complex integration of control logic. The intermediary approach reduces overall system complexity while enabling timely preventive action.
3Productivity
If congestion information from lower layer is fed to higher layer for congestion marking, then earlier detection and mitigation is enabled, but the protocol stack integration becomes more complex
Solution Approach 1:
The patent applies universality by designing the congestion information feedback mechanism to serve multiple functions: it enables early congestion detection, facilitates preventive marking actions, and maintains compatibility with existing ECN and PCN frameworks. This multi-functional approach increases network throughput while avoiding the need for entirely new complex integration protocols.
Solution Approach 2:
The feedback mechanism from data link layer to network layer creates a closed-loop system that improves productivity by enabling continuous monitoring and adaptive congestion control. The feedback information allows higher layers to adjust their marking behavior in real-time, optimizing network throughput without requiring complex rearchitecture of the protocol stack.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention provides a method and system wherein congestion information is taken from a lower level in the protocol stack and used for congestion marking in a higher layer. In particular, the method and system of the invention from a first aspect provide for congestion information in the lower layer to be generated downstream at a congested node, and to be fed back upstream to a source node. The source node then passes the congestion information, or information derived therefrom, up to the corresponding network element responsible for higher layer functions in the protocol stack, where higher level congestion marking can then be performed in dependence on the received lower level congestion information. Thus, congestion information from the lower layer which can typically be generated earlier before significant congestion occurs can be used in higher layer congestion control functions. Preferably the lower layer is the data link layer, and the higher layer is the network layer.