Method for distributing tag, computing route distance and implementing back-up switching in ring network by multiprotocol tag

Abstract

The invention relates to a method for making label allocation on ring network, calculating routing distance and implementing backup switching, and its kernel: separately allocting labels to node on ring network along clockwise and anticlockwise rings, respectively, and calculating routing distances between the node and lower-ring nodes, and based on the calculated results, respectively establishing corresponding relation between master and backup label exchange paths between the node and lower-ring node; when the master label exchange path fails, it switches the operating path to the backup label exchange path according to the corresponding relation. Thus, when adding / removing a node to / from a ring, the ring network can still operate; moreover, it is based on MPLS and transmits messages by the shortest path, thus raising transmitting efficiency; it can run failure detection protocol between adjacent nodes, thus increasing detecting and backup switching speeds; in addition, the invention supports inter-ring connection and connection with other MPLS networks.

Description

technical field [0001] The invention relates to the field of communication, in particular to a method for distributing labels, calculating route distances, and realizing backup switching by adopting multi-protocol label switching technology on a ring network. Background technique [0002] Multi-Protocol Label Switching (MPLS) technology introduces a label-based mechanism, which separates routing and forwarding. The label specifies the path for a packet to pass through the network, and data transmission is completed through the Label Switching Path (LSP). An LSP is unidirectional, and full-duplex services require two LSPs. [0003] At present, there are more and more applications on the network, and the network must be able to guarantee its service quality when a fault occurs. Therefore, each operator proposes a method for realizing link protection when a network failure occurs. [0004] The prior art one related to the present invention is to realize faulty link protection...

AI Technical Summary

Problems solved by technology

[0007] 2. Each section of optical fiber on the reciprocal double ring must work at the same rate. When the traffic has the characteristics of convergence, the utilization rate of optical fiber resources far away from the convergence point is low;

[0008] 3. The fault detection protocol needs to go around the ring, and the detection speed is slow;

[0009] 4. The length of the RPR header is relatively large, and the encapsulation efficiency is low;

[0010] 5. Cross-ring is not supported;

[0011] 6. Complex implementation, protocol processing requires hardware support, high cost

[0026] 1. The normal operation of the ring network depends on the only master node. If this node fails, the entire ring network will be unavailable;

[0027] 2. Only one direction on the ring network can forward packets, which is not the shortest path. The forwarding efficiency is low and bandwidth is wasted;

[0028] 3. The fault detection protocol needs to go around the ring, and the detection speed is slow;

[0029] 4. The forwarding based on the MAC header is slow;

[0030] 5. The existing technology can only be based on the Ethernet link layer

Images