OpenFlow Controller Topological Learning Across IP Networks

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The existing OPENFLOW network lacks efficient topological learning capabilities when interconnected with conventional IP networks, requiring manual configuration and resulting in low learning efficiency.

Innovation Solution

A topological learning method and apparatus that enables a controller to obtain and manage M OPENFLOW switch ports connected to a conventional IP network, determining the existence of a logical switch, creating and storing related link information, and using protocols like LLDP and BDDP to establish and update link information between OFS ports and the logical switch.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual configuration is used for topology of OPENFLOW network cross conventional IP network, then configuration can be completed, but topological learning efficiency is low

Engineering Contradiction:

Improvetopological learning efficiencyVSAvoidmanual configuration requirement

Core Design Contradiction:

ProductivityVSEase of operation

Solution Approach 1:

The system enables self-service topological learning by allowing the controller to automatically discover and learn the topology of conventional IP network segments through LLDP/BDDP protocol messages. The controller autonomously sends probe messages, receives responses, and builds topology information without requiring manual configuration, thereby improving productivity while eliminating manual operation requirements.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual configuration mechanisms with automated electronic discovery mechanisms. The controller uses LLDP/BDDP protocol messages to electronically probe and learn topology information, substituting the mechanical/manual process of configuration with an automated electronic system that efficiently discovers network topology through message exchange and automatic information building.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Extent of automation

If controller cannot perform topological learning through conventional IP network, then existing network architecture is maintained, but automation capability is limited

Engineering Contradiction:

Improvetopological learning automationVSAvoidcross-network learning capability

Core Design Contradiction:

Extent of automationVSAdaptability or versatility

Solution Approach 1:

The patent introduces an intermediary mechanism in the form of a logical switch that represents the conventional IP network segment. The controller interacts with this logical switch through standardized interfaces, enabling automated topological learning across the boundary between OPENFLOW and conventional IP networks. The logical switch acts as a mediator that translates and bridges the different network domains, allowing automation capability to extend into previously inaccessible areas.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The controller is enhanced with multi-functionality to perform both traditional OPENFLOW network management and new topological learning functions across conventional IP networks. The same controller infrastructure that manages OPENFLOW switches is extended to also discover, learn, and manage topology information in conventional IP network segments, making the system universally capable of handling multiple network types without requiring separate specialized systems.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Extent of automation

If conventional IP network is virtualized as logical switch, then automated configuration is enabled, but system complexity increases

Engineering Contradiction:

Improveautomated configuration capabilityVSAvoidlogical switch virtualization

Core Design Contradiction:

Extent of automationVSDevice complexity

Solution Approach 1:

The patent creates a virtual copy or representation of the conventional IP network segment in the form of a logical switch. This logical switch is not a physical device but a software-based abstraction that mirrors the functionality and topology of the physical network segment. By working with this simplified copy rather than the complex physical infrastructure, the system enables automated configuration while managing complexity through abstraction - the controller interacts with the simple logical model rather than the complex physical reality.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS10237166B2Topological learning method and apparatus for OPENFLOW network cross conventional IP network

Publication Date: 2019.03.19 HUAWEI TECH CO LTD

AI summary

A topological learning method and apparatus for an OPENFLOW network cross a conventional Internet Protocol (IP) network. The method includes obtaining, by a controller, M OPENFLOW switch (OFS) ports connected to a same conventional IP network, determining whether there is a logical switch corresponding to the conventional IP network, if the controller determines that there is no logical switch corresponding to the conventional IP network, creating and storing the information about the logical switch, where the information about the logical switch includes related information of the M OFS ports, and related information of each OFS port includes link information in a direction from the port to the logical switch and/or link information in a direction from the logical switch to the port, and managing, by the controller, the logical switch as a common OPENFLOW switch of an OPENFLOW network.