System and method for implementing fast layer 2 protection in passive optical networks

1. Technical Field of the Invention

The present invention generally relates to passive optical networks (“PONs”). More particularly, and not by way of any limitation, the present invention is directed to system and method for implementing fast layer 2 (“L2”) protection in PONs.

2. Description of Related Art

Wavelength Division Multiplex Passive Optical Networks (“WPONs”) rely on shared optical wavelength and unidirectional connections. Ethernet is used as the L2 technology in WPON. To prevent single point of failure, WPON employs dual-hub architecture; as a result, it is necessary to maintain a loop-free topology. However, the standard spanning tree cannot correctly function in such a network.

Currently, the solution to the foregoing problem is to utilize spanning tree protocols (“STP”) for providing both a loop-free network and protection. STP and the more recently developed rapid STP (“RSTP”) and multiple STP (“MSTP”) work efficiently in local area networks (LANS) and switched Ethernet networks for offering network protection at L2. This solution is deficient, however, in that bridges need bi-directional ports to exchange bridge protocol data units (“BPDUs”). Point-to-point connections are required to fully enable fast convergence. WPON is based on broadcast and shared optical bandwidth. Additionally, WPON nodes use asymmetric ports, i.e., some ports have only receivers, for efficient and cost effective data distribution. As a result, the broadcast and asymmetrical nature of the network prevents standard L2 protection schemes from functioning properly.

Therefore, what is needed is a system and method to provide L2 protection for networks that employ broadcast media and asymmetric traffic distribution.

One embodiment is a method for providing fast protection in a passive optical network (“PON”) including two central hubs and a plurality of network nodes connected to a first one of the central hubs by a first optical path and to a second one of the central hubs by a second optical path. The method comprises setting the first and second central hubs to master mode and slave mode, respectively, such that only the first central hub transmits traffic to and from the network nodes; detecting a failure in the first optical path; exchanging status information between the first and second central hubs; and if no failure has occurred in the second optical path, setting the first and second central hubs to slave mode and master mode, respectively, such that only the second central hub transmits traffic to and from the network nodes.

Another embodiment is a system for implementing a fast protection scheme in a passive optical network (“PON”) including two central hubs and a plurality of network nodes, wherein the network nodes are connected to a first of the central hubs by a first optical ring and connected to a second of the central hubs by a second optical ring. The system comprises means for setting the first central hub to master mode and the second central hub to slave mode such that only the first central hub transmits traffic to and from the network nodes via the first optical ring; means for detecting a failure in the first optical ring; means for exchanging control messages between the first and second central hubs regarding the ring status of the first and second optical rings; means for setting the second central hub to master mode and the first central hub to slave mode such that only the second central hub transmits traffic to and from the network nodes via the second optical ring if no failure has occurred in the second optical ring.

Another embodiment is a passive optical network (“PON”) for implementing a fast protection scheme. The PON comprises a first central hub connected to a plurality of network nodes via a first optical ring; a second central hub connected to the network nodes via second optical ring, the second central hub connected to the first central hub via a connection for enabling periodic exchange of ring status information between the hubs; wherein the first central hub is set to master mode and the second central hub is set to slave mode such that only the first central hub transmits traffic to and from the network nodes; wherein responsive to detection of a failure in the first optical ring, if no failure has occurred in the second optical ring, setting the first and second central hubs to slave mode and master mode, respectively, such that only the second central hub transmits traffic to and from the network nodes.

A more complete understanding of the present invention may be had by reference to the following Detailed Description when taken in conjunction with the accompanying drawings wherein:

FIG. 1 depicts a system block diagram of an exemplary PON for implementing an embodiment;

FIGS. 2A and 2B illustrate a response of the embodiment to a link failure in the PON of FIG. 1 ;

FIGS. 3A and 3B illustrate a response of an embodiment to a cable cut in the PON of FIG. 1 ;

FIG. 4 is a flowchart of the operation of a fast L2 protection scheme for a PON in accordance with one embodiment.

Embodiments of the invention will now be described with reference to various examples of how the invention can best be made and used. Like reference numerals are used throughout the description and several views of the drawings to indicate like or corresponding parts, wherein the various elements are not necessarily drawn to scale.

In general, a fast L2 protection scheme for a PON is proposed herein. Using a dual hub WPON as an example, in this approach, one of the hubs is in master mode, while the other is in slave mode. Two hubs using the proposed L2 protocol exchange underlying WPON topology status. During normal operation, only the master hub forwards traffic to and from the network nodes. When a failure occurs in a working path, or ring, of the network, the master hub detects the failure and immediately informs the slave hub to take over the traffic on the working links. At the same time, the master node stops forwarding the corresponding traffic to/from the network nodes to avoid duplication.

Referring now to FIG. 1 , illustrated therein is an exemplary WPON 100 for implementing an embodiment described herein. It should be recognized that, although the embodiment is illustrated as comprising a WPON, any type of PON may be used to implement the embodiments described herein. As shown in FIG. 1 , the WPON 100 comprises two hubs 102 , 104 , two optical paths, or rings 106 , 108 , and a plurality of network nodes, represented in the FIGS. by three nodes D 1 -D 3 , attached to both rings 106 , 108 . In one embodiment, the nodes D 1 -D 3 are digital subscriber line access multiplexers (“DSLAMs”). Each node D 1 -D 3 has one dedicated upstream channel on each ring 106 , 108 . Downstream traffic is delivered to the nodes D 1 -D 3 via the rings 106 , 108 , using shared wavelength. Traffic belonging to the same one of the nodes D 1 -D 3 is assigned to the same virtual local area network (“VLAN”), with the exception of broadcast TV (“BTV”) traffic, which uses specific broadcast VLANs seen by all of the nodes D 1 -D 3 .

Gigabit Ethernet (“GE”) ports (not shown) are installed at the hubs 102 , 104 , and nodes D 1 -D 3 to send and receive Ethernet frames. Thus, a loop-free active topology is required in the WPON 100 . However, as previously indicated, WPON uses, either fully or partially, shared wavelength for downstream traffic (i.e., from hub to node) and a dedicated wavelength for upstream traffic (i.e., from node to hub). Besides the BTV traffic, as previously indicated, the WPON 100 utilizes the asymmetric and unidirectional GE ports present at both the hubs 102 , 104 , and nodes D 1 -D 3 . As a result, spanning tree protocols, which require bidirectional links and supporting ports, cannot operate correctly on a network such as the WPON 100 .

In accordance with features of one embodiment, a fast L2 protection scheme is proposed in which each hub 102 , 104 , has a detection mechanism for detecting possible failures on one or both of the rings 106 , 108 . One of the hubs, e.g., the hub 102 , is in master mode and the other hub, e.g., the hub 104 , is in slave mode. By definition herein, the ring connected to the master hub is designated the working ring and the ring connected to the slave hub is designated the protection ring. The hubs 102 , 104 , periodically exchange underlying WPON topology status information using L2 control protocol via a connection 109 . In particular, the hubs 102 , 104 , exchange status information regarding the status of the rings 106 , 108 . Traffic from a provider network 110 is sent to both hubs 102 , 104 , using a 1+1 protection scheme. Each node D 1 -D 3 also sends the upstream traffic to both hubs 102 , 104 , using the 1+1 protection scheme. During normal operation, only the master hub 102 forwards the traffic to/from the nodes D 1 -D 3 via the working ring 106 . In contrast, the slave hub 104 drops, or filters out, the traffic that reaches its GE ports.

As will be described in detail below, each of the master and slave hubs 102 , 104 , determines whether to forward traffic on its corresponding ring 106 , 108 , respectively, based on the status of the rings. As previously noted, ring status information is periodically exchanged between the hubs 102 , 104 . The L2 protection method uses 1+1 for redundancy and blocks the slave hub port (i.e., filtering the traffic to and from the WPON 100 ) to keep the topology loop-free.

When a failure occurs in the working ring 106 , the master hub 102 detects the failure based on loss of the optical signal on both upstream and downstream links at the optical layer. A failure in the protection ring 108 is similarly detected by the slave hub 104 . Thus, the failure detection and location are fast. In response to detection of such a failure, the hubs 102 , 104 , exchange ring status information via L2 protocol control messages on the connection 109 as previously described and the protection process is initiated as will be described in greater detail below. In one embodiment, the L2 protocol control message comprises Ethernet/L2 frames. It should be noted that there is no timer involved in this process; rather, convergence time will typically depend primarily on failure detection and the control message processing time at the hubs 102 , 104 . Accordingly, the embodiments enable the performance of fast protection using the L2 protocol.

To better understand the operation of the proposed method, examples are presented in FIGS. 2A-2B and 3 A- 3 B, which respectively illustrate a case in which a failure occurs due to a fiber cut ( FIGS. 2A-2B ) and a cable cut ( FIGS. 3A-3B ). Referring to FIGS. 2A-2B , which illustrate the fiber cut case, when a fiber cut occurs in the working ring 106 , e.g., between nodes D 2 and D 3 , the master hub 102 detects the loss of downstream optical signals at the closing end of the ring 106 and the loss of the upstream signal from the node D 3 . As a result, the master hub 102 stops forwarding traffic via the working ring 106 and informs the slave hub 104 , via an L2 control protocol message, to start forwarding traffic via the protection ring 108 . As illustrated in FIG. 2B , as a result, the ports of the hub 102 are blocked, the ports of the hub 104 are unblocked, and traffic is routed to and from the network 110 via the hub 104 . Additionally, traffic is routed between the hub 104 and the nodes D 1 -D 3 via the ring 108 .

Referring now to FIGS. 3A and 3B , which illustrate the cable cut case, assuming a cable cut occurs between the nodes D 2 and D 3 , thereby affecting both rings 106 , 108 , both hubs 102 , 104 , need to forward the correct VLAN traffic via both rings 106 , 108 . In particular, when the master hub 102 detects the loss of downstream optical signals at the closing end of the ring 106 and the loss of the upstream signal from the node D 3 and the slave hub 104 detects the loss of upstream optical signals from the nodes D 1 and D 2 , the hubs exchange L2 control protocol messages to advise the other regarding the status of their respective ring. As a result, as illustrated in FIG. 3B , the ports of the hub 102 are partially blocked such that the VLAN corresponding to the node D 3 is dropped. Similarly, the ports of the hub 104 are partially blocked such that the VLAN corresponding to nodes D 1 and D 2 is dropped. As a result, traffic to and from the node D 3 is routed via the hub 104 and the ring 108 , while traffic to and from the nodes D 1 and D 2 is routed via the hub 102 and the ring 106 .

FIG. 4 is a flowchart of the operation of an embodiment for providing fast L2 protection in a PON, such as the WPON 100 . Initially, one of two hubs is set to master mode and the other is set to slave mode (block 400 ). The hubs monitor the status of the rings and periodically exchange status information using, for example, Ethernet/L2 frames (block 402 ). Responsive to detection of a failure in the working ring (block 404 ), if the protection ring has not failed (block 406 ), the hub currently set to master mode is set to slave mode, while the hub currently set to slave mode is set to master mode (block 408 ) and traffic is blocked from transmission via the failed ring (i.e., the previous working ring) and is transmitted via the new working ring (i.e., the ring connected to the new master hub) (block 410 ).

If failures are detected in both the working ring (block 404 ) and the protection ring (block 406 ) (i.e., if the cable is cut), the master hub is partially blocked such that only traffic to and from nodes still connected to the master hub is transmitted via the working ring (block 412 ). Similarly, the slave hub is partially blocked such that only traffic to and from nodes still connected to the slave hub is transmitted via the slave ring (block 414 ).

It should be noted that, although the embodiments described herein are described with reference to optical rings, other types of optical path configurations that are known or may become known may also be employed.

An advantage of the embodiments described herein is that they provide L2 protections for both link failure and hub node failure. Moreover, they provide instant switchover to the protection node/ring upon detection of link failure; no timer is required during failure handling.

Another advantage of the embodiments described herein is that control information is only exchanged between hubs. Light weight local operation imposes very light overhead on the network. Additionally, no new hardware or software is required at the nodes; the protocol runs only on the hubs.

It is believed that the operation and construction of the present invention will be apparent from the Detailed Description set forth above. While the exemplary embodiments of the invention shown and described have been characterized as being preferred, it should be readily understood that various changes and modifications could be made therein without departing from the scope of the present invention as set forth in the following claims.