Software Defined Network Node Control Traffic Relaying

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Centralizing the network forwarding plane in software defined networks (SDNs) can lead to delays in network reconfiguration, especially when integrating new nodes or recovering from failed control connections, as all decisions must be made and transmitted by a central controller, which can be cumbersome for unpredictable or sudden events.

Innovation Solution

Implementing network-node-initiated exchange of control messages between nodes, allowing for relaying of control traffic and data flow table manipulation, enabling seamless integration of new nodes and recovery from failed connections without relying solely on central controller decisions, by using existing end-to-end control connections and detecting link layer failures for quick path reestablishment.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If a central controller is used to manage all network nodes, then network control is centralized and streamlined, but network reconfiguration delay increases when integrating new nodes or recovering from failures

Engineering Contradiction:
Improvecentralized controlVSAvoidnetwork reconfiguration delay
Core Design Contradiction:
Extent of automationVSLoss of time

Solution Approach 1:

The patent segments the centralized control function by introducing hierarchical controllers (regional controllers) that can autonomously manage subsets of network nodes. This segmentation allows local reconfiguration decisions to be made without waiting for the central controller, reducing reconfiguration delay while maintaining overall centralized architecture.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediary controllers that act as mediators between the central controller and network nodes. These intermediaries can handle control messages and reconfiguration decisions locally, reducing the burden on the central controller and accelerating network reconfiguration for local events.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Device complexity

If all control decisions are made by a central controller, then control plane protocols are streamlined, but the burden on establishing new control connections increases

Engineering Contradiction:
Improvecontrol plane protocol complexityVSAvoidcontrol connection establishment
Core Design Contradiction:
Device complexityVSEase of operation

Solution Approach 1:

The patent implements preliminary action by having controllers pre-establish control connections and maintain connection pools before they are needed. When new nodes are integrated or failures occur, pre-established connections can be quickly activated without requiring new connection establishment, reducing the operational burden.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent enables self-service by allowing network nodes and controllers to autonomously establish and manage their own control connections without requiring manual pre-configuration. The system automatically discovers available connections and establishes control paths as needed.

Inventive Principle:
Principle #25Self-service

3Device complexity

If control connections require preconfiguration or existing paths, then controller management is simplified, but integration of new network nodes becomes cumbersome

Engineering Contradiction:
Improvecontroller management complexityVSAvoidnew node integration
Core Design Contradiction:
Device complexityVSEase of manufacture

Solution Approach 1:

The patent enables new network nodes to self-configure and automatically establish control connections with appropriate controllers without manual pre-configuration. Nodes can discover available controllers and autonomously set up control paths, simplifying integration while maintaining manageable controller complexity.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent employs temporary or ephemeral control connection objects that can be quickly created and destroyed as nodes are integrated or removed. These short-lived connection objects allow flexible node integration without requiring permanent pre-configuration infrastructure.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

4Extent of automation

If the central controller handles all reconfiguration decisions, then centralized control is maintained, but network efficiency decreases during sudden events or changes

Engineering Contradiction:
Improvecentralized control decision-makingVSAvoidnetwork reconfiguration speed
Core Design Contradiction:
Extent of automationVSProductivity

Solution Approach 1:

The patent segments the decision-making authority by enabling regional controllers and network nodes to make autonomous reconfiguration decisions for local events. This segmentation allows immediate response to sudden changes without waiting for central controller processing, improving network productivity while maintaining centralized oversight for major decisions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic control where the degree of centralization can adapt based on the situation. For routine or local events, decentralized autonomous decision-making is enabled for faster response. For significant network-wide changes, centralized control is activated. This dynamic approach optimizes both centralized control maintenance and reconfiguration speed.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3350963B1Control traffic in software defined networks
Publication Date: 2021.08.04 HUAWEI TECH CO LTD
  • EP3350963B1 patent drawingFigure 1~2
  • EP3350963B1 patent drawingFigure 3~4
  • EP3350963B1 patent drawingFigure 5~6

AI summary

Provided is a first network node having a data flow table identifying forwarding rules to be applied to incoming data packets, wherein the first network node is configured to allow for manipulation of the data flow table via control messages received via a control connection to a controller, wherein the first network node is further configured to determine an occurrence of a network event or compliance with a condition specified by the controller, wherein the first network node is further configured: to establish a control connection to a second network node in case that the network event occurs or the condition is met and to exchange control messages with the second network node via the control connection.