Bidirectional Encapsulation Translation for SPBm Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional network edge virtualization encapsulation techniques, such as VXLAN and NVGRE, rely on complex IP multicast mechanisms in L3 networks, which can be inefficient and cumbersome, especially when handling multicast traffic, and often require modifications to customer data frames.
Innovation Solution
The method involves bidirectional translation of edge network virtualization encapsulation to core network virtualization encapsulation using MAC-in-MAC encapsulation over a shortest path bridging (SPBm) network, which simplifies multicast mechanisms and reduces header size, allowing efficient forwarding of unicast and multicast traffic without altering the customer data frame.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional VXLAN or NVGRE encapsulation techniques are used to extend L2 networks across L3 networks, then network virtualization functionality is achieved, but the system complexity increases due to reliance on complex IP multicast mechanisms in L3 networks
Solution Approach 1:
The patent introduces SPBm (Shortest Path Bridging MAC) as an intermediary L2 multicast mechanism to replace complex L3 IP multicast. The SPBm network acts as a mediator that provides efficient multicast forwarding between edge networks, eliminating the need for customer networks to implement complex IP multicast protocols while maintaining network virtualization functionality.
Solution Approach 2:
The patent extracts the multicast functionality from the L3 IP network layer and relocates it to the L2 network layer using SPBm. This separation allows the L3 network to be simplified while maintaining multicast capabilities at the L2 layer where they are more efficiently handled by SPBm's integrated routing and switching architecture.
2Adaptability or versatility
If conventional VXLAN or NVGRE encapsulation techniques are used, then L2 network extension is achieved, but header overhead increases due to addition of encapsulation headers
Solution Approach 1:
The patent merges the L2 forwarding information with the L3 routing information in the SPBm header. By combining these functions into a single header structure, the patent eliminates the need for separate VXLAN or NVGRE encapsulation headers, thereby reducing overall header overhead while maintaining L2 network extension capability across the L3 SPBm network.
3Adaptability or versatility
If conventional VXLAN or NVGRE techniques are used with L3 multicast, then network virtualization is achieved, but processing efficiency decreases when handling multicast traffic
Solution Approach 1:
The patent replaces the complex L3 IP multicast mechanism with the L2 SPBm multicast mechanism. This substitution leverages the simpler and more efficient L2 forwarding plane of SPBm switches, which can handle multicast traffic more efficiently than L3 routers running IP multicast protocols, thereby improving multicast traffic processing efficiency while maintaining network virtualization.
Data Source
AI summary
A method, system and computer readable medium for bidirectional translation of edge network virtualization encapsulation and core network virtualization encapsulation are described. The method can include receiving a customer data frame and removing an edge network virtualization encapsulation header (e.g., VXLAN, NVGRE or VLAN header) and adding a core network virtualization encapsulation header (e.g., a MAC-in-MAC header). The customer data frame can be forwarded in a core network (e.g., an SPBm network) utilizing the core network virtualization encapsulation header, which can be translated back to an edge network virtualization encapsulation header at a receiving switch.