0. Preface
There are many different link resources in intelligent Optical network, such as fiber link, wavelength link, various STM link and VC layer link in SDH, and various OTM layer link and och layer link in otn, etc. These links are called Sub dot (SNP) links in Ason, for the scalability of routing protocols and routing diversity, several different SNP links can be combined into sub-dot pool (SNPP) links (These SNP links are called data links in Gmpls, and these SNPP links are called Business Volume Engineering (TE) Link). In Ason, these SNPP link resources are managed by the Link Resource Manager (LRM), which allocates or frees the SNP link connections based on the request of the connection controller. But before you manage and control these SNPP link resources, you first need to discover these link resources and their control entities, which is the automated discovery technology, which is implemented by Discovery Agents (DA), Terminals and Adapters (TAP), and link explorer.
The automatic discovery technique includes the discovery of adjacency on the physical media layer, the discovery of layer adjacency and the discovery of logical adjacency of the control entity. The adjacency discovery on the physical media layer is used to confirm the physical connectivity between the two ports of the adjacent network element on the physical media layer (through optical fibres or other physical media). The layer adjacency discovery is to confirm (or infer from the relationship between SNP and TCP) the correlation of a pair of SNP between two adjacent network elements, thus forming a SNP link connection on a particular layer network. As long as the path supporting the SNP link is valid, the correlation obtained through the layer adjacency discovery is valid. The logical adjacency discovery process of control entities is to establish the logical adjacency relationship between the two control entities controlling these SNP link resources.
In this paper, a method of realizing neighbor Discovery and Link resource discovery based on optical Monitoring channel (0SC) in Intelligent Optical Network is proposed, including the discovery of adjacent control entities ' automatic discovery and SNPP (SNP) link.
1, the principle of automatic discovery
In order to manage connections in a layer network, in the control plane, the underlying transfer plane resources are represented by some entities, as shown in Figure 1, which is the relationship between the different transfer plane resources described in the ITU-T g.805, A view of the entities that are convenient for network management to represent these resources and the transport plane resources seen by the control plane. From this diagram and the ITU-T g.8080, it is known that the routing function in Intelligent Optical Network is to understand the transport plane topology by SNPP link connection, before the SNPP link connection is established, the lower level transport plane topology must be established, that is, the link connection between the connection point (CP). The connection between SNP and the connection between CP can be realized by artificial configuration through network management system, and can be discovered automatically by using different mechanisms, such as using the test signal in the customer layer net load or using the service layer path trail byte to discover these link connections.
Fig. 1 relationship between different structural entities in the transport plane, the management plane and the control plane
The automatic discovery in Intelligent optical Network is mainly by Discovery Agent (DA), terminal and adapter (TAP) and Link Resource Manager (LRM), as shown in Figure 2, where the agent is found to be fully operational for the transport plane's namespace (that is, for CP), to maintain the connection between the CP and the CP ; The terminals and adapters operate on the namespace of the transport plane and on the control plane's name space. is to maintain the binding relationship between SNP and CP, while link Explorer is completely operational for the namespace of the control plane, managing the link between SNP and SNP, Routing controllers also select routes for these SNPP links (or SNP links). The goal of automatic discovery is to eventually obtain SNPP (SNP) link connections between adjacent nodes.
Fig. 2 Interaction between different control components when automatic discovery
As mentioned above, to obtain the SNPP (SNP) Link connection, it is necessary to obtain the binding relationship between SNP and CP of two neighboring nodes, and the connection between the CP of the neighboring nodes. As we know from Figure 2, SNP link connections can be obtained in two ways, the first of which is to use the path Trail ID to discover the connection between Peer-to-peer TCP (Tcp3s and tcp3r in Figure 2), which is to load discovery messages in the path trail byte of the service layer, and transmit discovery messages to neighboring nodes. The neighboring node then transmits the response message from the control channel, obtains the link connection between the tcp3s and the tcp3r, and then, according to the relationship between the local configuration and the adaptation function and the path terminal function, obtains the link connection between the CP and the neighboring node CP. According to the binding relationship between SNP and CP which is configured and maintained by this node, the link connection between SNP and neighboring node is obtained.
Another approach is to use test signals to discover the connection between Peer-to-peer TCP (the Tcp1s and tcp1r in Figure 2, that is, loading discovery messages in the net load of customer layer signals, transferring discovery messages to neighboring nodes, and then transmitting discovery response messages from the control channel). The link connection between the Tcp1s and the TCP1R is obtained, and the link connection between the local CP and the neighboring CP is obtained based on the previously established matrix connection (to connect the test signal to the desired CP). According to the binding relationship between SNP and CP which is configured and maintained by this node, the link connection between SNP and neighboring node is obtained. Obviously this method can only be used if the link connection does not host the customer tier business.
2, the use of optical monitoring channel automatic discovery process
The automatic discovery process in intelligent optical networks can be divided into several stages, namely, the discovery of resources in local nodes, the discovery of neighboring nodes and physical media layer, the discovery of layer adjacency and the interaction of business attribute information. The resource discovery of the local node is the control entity in this node first to determine the local Snp-id, Snpp-id, Tcp-id, Cp-id, SNP-CP binding relationship, TCP-CP binding relationship, etc.
From the above discovery mechanism can be seen, whether it is to send the client layer test message or use the Service Layer path trail byte method, all need to find the physical media layer adjacency and the adjacent node address on the basis of the discovery of the adjacent relationships on each layer. Figure 3 shows the test method used to discover the address information of the neighboring nodes. Two Intelligent Optical network nodes A and z, connected by a fiber, for the sake of simplification, did not draw the relevant components of the control plane, nor consider the fiber-level link connection and various VC layer link connections.
(1) Adjacent discovery of physical media layer and discovery of neighboring nodes
As shown in Figure 3, Node A sends the test message, which should include the IP address and name of Node A, the address and name of the Discovery Agent (DA) in Node A, the address and name of the terminal and adapter (TAP), and the local wavelength port (Tcp-id). The message uses PPP/HDLC frame format, modulation to the wavelength λ1, from the sending module to the wavelength of λ1 optical switch, and cross connected to the specified port, the port's wavelength signal by multiplexer, optical amplifier, optical fiber, etc. to Node Z. When the test message is received, node Z obtains the IP address and name of Node A, the address (Da-a-id) and name of the agent found in Node A, the address (Tap-a-id) and name of the terminal and adapter, and the local wavelength port identifier (TCP-ID) used by Node A. When a discovery response message is received, both Node A and node Z are identified as adjacency nodes.
Fig. 3 schematic diagram of adjacent node discovery using Test signal
