Masked Ethernet Addresses for TRILL Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing transport networks, such as TRILL networks, face challenges in extending routing and switching functionality due to Ethernet encapsulation, which reduces routing capabilities and increases traffic delays, and introducing new routing protocols requires substantial hardware and software updates, making it costly and time-consuming.
Innovation Solution
The solution involves masking a portion of information within Ethernet address headers using an address mask for TRILL headers, allowing additional bits to be used for metadata or routing operations without increasing packet size or burdening RBridges with complex implementation, enabling features like ECMP and flow path identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If Ethernet encapsulation is used in transport networks, then data transport is simplified and accelerated, but routing functionality is reduced and traffic delays increase
Solution Approach 1:
The Media Access Control (MAC) address is segmented into two functional parts: a routed portion (typically the first 3 bytes) used for transport network routing decisions, and a local portion (remaining bytes) preserved for end-device identification. This segmentation allows RBridges to perform routing based on the routed portion while maintaining compatibility with existing Ethernet addressing schemes, thus preserving routing functionality while benefiting from Ethernet encapsulation performance.
2Adaptability or versatility
If deep packet inspection is implemented to add routing features, then additional routing capabilities are achieved, but network traffic slows down
Solution Approach 1:
The routed portion of the MAC address is extracted and used for routing decisions at the ingress RBridge before packet forwarding begins. This preliminary extraction of routing information from the MAC address eliminates the need for deep packet inspection at intermediate RBridges, as routing decisions can be made based solely on the pre-positioned routed address portion, thereby maintaining high forwarding speed while achieving enhanced routing capabilities.
3Adaptability or versatility
If a new routing protocol is introduced to extend functionality, then routing capabilities are enhanced, but hardware and software updates are required across all nodes
Solution Approach 1:
The existing MAC address field in Ethernet frames is made multi-functional by designating a portion of it (the routed portion) for transport network routing purposes while preserving the local portion for end-device identification. This universal reuse of the existing MAC address structure allows enhanced routing capabilities without requiring new protocol fields or hardware modifications, as the same MAC address infrastructure serves both traditional Ethernet addressing and TRILL routing functions.
4Measurement precision
If all MAC address bits are used for addressing, then addressing precision is maximized, but additional routing information cannot be included
Solution Approach 1:
The 48-bit MAC address is segmented into a routed portion (first 3 bytes, 24 bits) and a local portion (remaining 24 bits). The routed portion contains precision addressing information for identifying destination RBridges in the transport network, while the local portion preserves precision identification of end devices. This segmentation allows both precise addressing and inclusion of routing information to coexist within the same address structure.
Data Source
Figure 1A~1B
Figure 2
Figure 3
AI summary
A computer-implemented method for packet switching in a transport network is described. The method comprises receiving a data packet at a first data switching device, the data packet having an existing header and being received from a customer network, the first data switching device being an edge node of a transport network. Encapsulating the data packet using a transport network header, the transport network header having a data structure that includes a transport device address space and a virtual local area network, VLAN, indicator space. Encapsulating the data packet includes setting a first portion of bits within the transport device address space, such that the first portion of bits indicates a data switching device address that is a node within the transport network. Encapsulating the data packet includes setting a second portion of bits within the transport device address space, such that the second portion of bits indicates a flow identifier. Selecting a forwarding path to a next-hop node of the transport network by using the flow identifier from the transport device address space as computational input for selecting the forwarding path. Forwarding the data packet to the next-hop node in the transport network using the selected forwarding path.