Multi-layer Implementation of optical network recovery policies

Restoration Technology is the most effective means for the mesh optical network to effectively cope with network faults and ensure service transmission in the network. Recovery techniques can be categorized and implemented in a variety of ways from different perspectives.

Link-level and channel-level

Based on the recovery granularity, optical network fault recovery can be divided into link-level recovery and channel-level recovery.

Link-level recovery is based on links. It mainly finds alternative routes for faulty links and implements unified recovery for all services on the faulty links. This recovery method mainly uses a local detection mechanism to re-Select Services that pass through the fault section. When a link fails, a new path is established between the two endpoints of the faulty link. This path and the original service path are not affected by the fault to form a recovery path.

Channel-level recovery is based on the business. The Restoration route and wavelength are calculated between the source and target nodes, and there is no algorithm-based Dynamic correlation between the recovered services. Compared with Link-level recovery, the restored optical channels established by channel-level recovery do not have to continue to maintain the parts that have not failed in the original working channel.

Pre-computing and real computing

The recovery optical channels can be divided into pre-calculation recovery and real-time computing recovery.

In the pre-calculation policy, routes for restoring optical channels are calculated offline based on certain calculation policies before a fault occurs. In the calculation, all possible faults must be taken into account and the corresponding recovery optical channels must be calculated. Then, information about the recovery channels should be distributed to all nodes, enables each node to immediately recover after detecting a fault, so as to achieve rapid fault recovery.

A Real-time computing policy is used to calculate the recovery path after a fault occurs. Therefore, it is necessary to locate the fault in real time, detect idle resources, and then calculate the recovery route.

Compared with the pre-computing policy, the real-time computing policy recovery speed is slow. However, the real-time Fault Locating and real-time resource discovery policies allow route recovery computing to take into account the current network topology and available resources, therefore, it has a high recovery success rate and network resource utilization.

Centralized and distributed

Based on the difference in the control mechanism of the computer-based routing and restoration implementation, the restoration technology can be divided into two methods: centralized recovery and distributed recovery.

The biggest difference between centralized recovery and distributed recovery lies in the different ways in which route recovery is calculated, or the difference in the information that each node has when calculating and restoring a route, restoring the executors of route computing or restoring the storage location of route computing results is not a sign of distinguishing the two recovery methods.

In a centralized restoration algorithm, the compute node for routing restoration can be the central control node of the network or the source/destination node of the damaged service, the requirement is that the node must master the link status information of the entire network, including the topology information, current resource usage information, route table, and failure location.
In a distributed policy, the compute process for Route recovery is usually started by the source/destination node of the Service or the end node of the faulty link. In distributed recovery, each node in the network does not need to obtain the link status information of the entire network. You only need to compute the route Recovery Information Based on the status information of the adjacent link. When a network failure occurs, the source node starts optical channels for recovery. First, the source node will send a recovery request (route search information) to the network broadcast service. When the Service's local node receives a recovery request from the service source end, the reply message is reversed to the source node along the route that receives the broadcast information, and the intermediate node is configured to establish a recovery optical channel. When the Response Information arrives at the business source end, the source node sends a confirmation message to the local end, and the business recovery route is established completely.

In addition, with the increasing development of communication services, the establishment of business level agreement management and business topology model will also become a key technology in network protection and business management.

Related Articles]

• Ultra High Speed: optical switching/Optical routing creates all Optical Networks
  

• Research on hierarchical routing technology of Automatically Switched Optical Networks
  

• Detailed process of evolution of optical network infrastructure