Architecture of ASON routing technology (1)

1. Preface

The emergence of automatic switching optical network ASON is a historic breakthrough in the development of the transmission network. ASON, also known as intelligent optical network, is a network that uses an independent control plane to dynamically Configure connection management. The core technologies of the ASON architecture include the signaling protocol, routing protocol, and link resource management. The signaling protocol is used to establish, maintain, and delete distributed connections. The routing protocol provides routing services for establishing connections. Link resource management is used for Link Management, verification and maintenance of control channels and transport links are provided.

This article mainly discusses ASON routing technology. First, it introduces the architecture of ASON routing technology, including the technical requirements of ASON routing and the routing information distribution topology. Then we discuss the differences between ASON routing technology and IP network routing. Next, we will introduce how ietf gmpls expands the IP routing protocol. Finally, the progress of OIF in NNI routing protocol is introduced.

2. ASON Routing Architecture

The ITU-T G.7715 defines the structure and requirements for establishing SC and SPC connection routing functions in ASON networks. The main content includes functional components such as ASON Routing Structure, path selection, route attributes, abstract information, and status chart transfer. G.7715 aims to provide a protocol-independent method to describe the routing technology used for ASON. Routing messages are transmitted through the data communication network DCN. G.7712 standardizes a possible implementation method of DCN. To provide the routing service, you need to know the network resources in advance. These resources can be manually configured or automatically discovered.

The ASON routing architecture supports different routing methods defined by G.8080, such as hierarchical routing, hop-by-hop routing, and source routing. This structure also abstracts different routing information expressions, such as link states and distance vectors. The ASON routing architecture is divided into multiple routing domains in the network and is used after network resources are allocated.

(1) Basic Concepts

Carriers can divide networks based on specific policies based on geography, management scope, and technology. The carrier can regard the subdivided network as composed of different routing domains to provide routing services. The routing domain provides the abstraction of route information, which makes the representation of route information scalable. A routing domain provides services such as channel computing through the routing executor RP). The routing executor is a consortium of the routing controller RC. Each RP is responsible for controlling a routing domain. The RP supports channel computing in the routing domain where it provides routing services, and is consistent with the source route, hierarchical route, and hop-by-hop route example defined by G.8080 ). The channel computing function supported by RP is based on the information type provided by the routing information database.

A route domain can be classified. In a route level, each route domain is associated with an independent RP. Each layer of a route level can use RP that supports different routing modes. The implementation of RP can be based on a distributed routing controller. RC provides a routing service interface, which is a service access point defined by RP. RC is responsible for coordination and distribution of route information. The RC service interface provides the routing service at a given level of NNI reference point. Different RC instances may be subject to different policies because they provide different services in different organizations. Policies can be executed through different mechanisms, such as using different protocols.

The implementation of RC can be a group of distributed entities, which are called a route control domain RCD ). RCD is an abstract entity that hides the internal details of the routing control domain and provides interfaces with the same features as the RC distribution interface. The routing information attributes exchanged between RCD include the common semantics of the routing information exchanged between RC distribution interfaces, and allow different expressions in each domain. The implementation of RCD depends on the specific implementation method.

The relationship between RA, RP, RC, and RCD is shown in Figure 1.

Figure 1 Relationship between RA, RP, RC, and RCD

1 shows that the routing domain contains the routing domain and recursively defines the continuous hierarchical routing level. An independent RP is associated with a routing domain. In this way, the RP itself is implemented by the distributed RC, RC1 comes from RPRA, and RC2 comes from RPRA.1 and RPRA.2. It can be found that the features of the RCD distribution interface and the RC distribution interface are consistent.

(2) routing architecture and functional components

The routing architecture includes protocol-independent components such as link resource manager LRM) and RC, and protocol-related components such as protocol controller PC ). RC processes the abstract information used for routing. The PC processes protocol-related messages based on the reference points that pass through the information, such as E-NNI, I-NNI, and passes routing primitives to RC. Figure 2 shows an example of the routing feature.

Figure 2 routing feature example

RC): RC functions include exchanging routing information with the RC on the peer end, and replying to the route query path selection through operations on the route information data packets ). RC is not related to the Protocol.

Route information database RDB): RDB stores local topology, network topology, accessibility, and other information obtained through route information exchange, as well as configuration information. RDB can contain the routing information of multiple routing domains. RC can be used to access a view relational database of RDB. A view is a virtual table. It has no storage space, but is exported from the tables actually stored in the database. The usage of the view is the same as that of the physical table. Different users can see different formats of the same data and use the authorization mode to control users' access to sensitive data ). The dotted box in Figure 2 shows this relationship. RDB is protocol-independent. Because RDB can contain the routing information of multiple routing domains, that is, it may be a multi-layer network), the RC connected to RDB may share the routing information. 3.

Link resource manager (LRM): LRM provides all SNPP link information to RC, and notifies RC of any state changes to the link resources it controls.

Protocol controller PC): a PC converts a routing primitive into a protocol message of a specific routing protocol, so it is related to the Protocol. The PC also processes protocol-related control information for route information exchange.

Figure 3 Relationship Between RDB and multiple RC/routing Domains

(3) classification of routing Domains

Figure 4 shows an example of a routing domain. The parent of the High-level routing domain) RA contains the lower-layer routing domain sub-) RA.1, RA.2, and RA.3. Similarly, RA.1 and RA.2 further include the route domains RA.1.x and RA.2.x.

Figure 4 routing domain classification example