Carrier Ethernet Technologies for the NGN Era

2.1 Ethernet OAM technology

Ordinary Carrier Ethernet generally does not include standard maintenance and management functions. Instead, fault troubleshooting and other tasks are performed using the ping command and/or proprietary tools from switch vendors. However, the use of commands on the Internet protocol level (layer 3 (L3)) and higher levels means that normality checking using test frames on the Ethernet level (layer 2 (L2)) is not performed. Furthermore, when a network has been built using switches from multiple vendors, maintenance personnel have no choice but to use the tools provided for each type of switch.

By contrast, next-generation Ethernet will make use of standard management functions provided by Ethernet OAM technology, enabling maintainability and operability to be improved (Table 1). Ethernet OAM technology, which was standardized as ITU-T Y.1731 in May 2006 and as IEEE802.1ag in September 2007, features a function set necessary for maintenance and operation work (constant monitoring of each virtual local area network, continuity checking, checking of frame transfer paths, and measurement of transmission delay). Configuration functions at the management level can also be used to provide network security by preventing the transfer paths of OAM frames transmitted from a user terminal from being detected.

Table 1. Comparison of maintenance and operation technologies.

2.2 ERP technology

Transmission paths can generally be made redundant by establishing a ring configuration in the relay network (Fig. 1). If a fault occurs within a ring network, processing for switching the frame transfer path must be performed while preventing frames from falling into an infinite loop that can put pressure on the transmission-path bandwidth. However, this switching can take several seconds to complete when using a spanning tree protocol, a general-purpose L2 path control technology, and such a long break in communications presents a technical problem to be solved.

Fig. 1. Example of ERP technology in operation.

In response to this problem, we have developed ERP technology having the following features toward next-generation Ethernet.

• Loop prevention at the time of path switching

• High-speed path switching

Taking a typical ring configuration (Fig. 1) as an example, we assume that a fault occurs in switch A (1). Then, as adjacent switches B and C detect that fault, the ports on those switches that connect to switch A are blocked (2). Next, other switches are notified of the fault (3), and the port on switch D, which is normally blocked, is now released, thereby constructing a new path for frame transfer (4). Finally, the MAC (media access control) address lookup table is initialized at all switches on the ring, and frame transfer processing on the new path begins (5).

The port-blocking processing of step (2) achieves switch processing without generating loops, and the simple process of blocking and releasing ports in steps (2)–(4) enables a path to be reconstructed and high-speed path switching to be performed.

The development of this technology has prompted standardization efforts in ITU-T SG15 (International Telecommunication Union, Telecommunication Standardization Sector Study Group 15) resulting in the approval of Recommendation G.8032/Y.1344 [1], [2] in June 2008.

2.3 Access-network redundancy technology

Redundancy technology in the access network prepares two systems (active system and backup system) each consisting of an optical network unit and optical subscriber unit in the access network (Fig. 2). When a fault is detected, the active and backup systems are switched over, which improves the reliability of the access network.

Fig. 2. Network configuration using access-network redundancy technology.

In this way, ERP technology and access-network redundancy technology combine to achieve a highly reliable network having redundancy functions.