Forced MAC Learning in Bridged Ethernet Networks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current Ethernet technologies are connection-less, making it difficult to apply traditional path provisioning techniques in transport networks, which require connection-oriented features, and existing solutions like GMPLS-controlled Ethernet Label Switching add complex protocol stacks, increasing costs.

Innovation Solution

The implementation of a constrained-based source-routing method using Ethernet MAC learning for path provisioning in bridged Ethernet networks, employing a lightweight protocol stack to establish connection-oriented paths with a path learning frame mechanism that constructs and forwards lists of links with destination and source addresses, and VLAN IDs through intermediate bridge nodes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If traditional path provisioning techniques are applied in Ethernet networks, then connection-oriented features are achieved, but device complexity increases due to complex protocol stacks

Engineering Contradiction:
Improveconnection-oriented featuresVSAvoidprotocol stack complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent uses MAC learning frames that copy the structure of existing Ethernet frames while incorporating path provisioning information. Instead of implementing entirely new complex protocols, the invention embeds path setup capabilities within standard Ethernet MAC frame structures, allowing bridges to learn and forward paths using extended MAC learning functionality rather than complex new protocol stacks

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent makes the existing MAC learning mechanism multi-functional by enabling it to perform both traditional MAC address learning and path provisioning functions. The MAC learning frame serves dual purposes: maintaining standard Ethernet bridging operations while simultaneously establishing connection-oriented paths, eliminating the need for separate complex protocol stacks

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Productivity

If GMPLS-controlled Ethernet Label Switching is implemented, then path provisioning capability is improved, but device complexity and cost increase

Engineering Contradiction:
Improvepath provisioning capabilityVSAvoidprotocol stack complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent uses lightweight, simple MAC learning frames instead of complex GMPLS protocol structures. These frames are simple Ethernet-based structures that can be processed by standard bridge hardware, avoiding the need for expensive complex protocol stacks while maintaining path provisioning functionality

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

Solution Approach 2:

The patent replaces the mechanical complexity of GMPLS protocol stacks with a simpler Ethernet MAC learning mechanism. Instead of implementing complex signaling protocols and state machines, the invention uses extended MAC learning frames that leverage existing Ethernet bridge functionality, substituting complex control plane mechanisms with simpler data plane operations

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS8687519B2Forced medium access control (MAC) learning in bridged ethernet networks
Publication Date: 2014.04.01 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)
  • US8687519B2 patent drawing
  • US8687519B2 patent drawing
  • US8687519B2 patent drawing

AI summary

A system, method, and Ethernet bridge utilizing forced MAC learning to set up a path from an originating node, PON (23) to one or more terminating nodes, PTN (12) through intermediate bridge nodes (14, 21). In the point-to-point case, a Previous Hop Bridge (PHB) list defines the links in the path. Each link includes a destination address, source address, port number, and VID. A path learning frame (13, 19, 22) containing the list is sent in reverse order from the PTN through the intermediate bridge nodes to the PON. Each node in the path saves the received frame's header information as learned path information using standard MAC learning functionality, removes the node's own information from the list, and forwards the frame with the remainder of the list to the next node using the port specified by the port number for that link. When the frame arrives at the PON, all nodes have saved the learned path information.