Ethernet protection system

CA 02446239 2010-01-05

02/099676

PCT/USO2/17921

ETHERNET PROTECTION SYSTEM

BACKGROUND

1. Field of the Invention

The present invention is related to the field of data communications.

More specifically, the invention relates to protecting a network connection between two Ethernet devices in the network. Such connections may include, for example, a connection between a SONET network element and an Ethernet switch.

2. Description of the Related Art

Several techniques exist for protecting network connections, such as in a Local Area Network (LAN) or a Wide Area Network (WAN). Two popular protection schemes are the spanning tree protocol, which is documented in the IEEE 802.ID standard, and trunking, which is documented in the IEEE 802.3ad standard.

The spanning tree algorithm allows Ethernet devices, such as bridges and switches, in an Ethernet network to dynamically create a loop-fiee set of paths. Loops in an Ethernet network may result in forwarding loops, in which broadcast and multicast traffic circulates endlessly and continues to grow as new broadcast and multicast traffic is transmitted. Eventually the traffic rate gets so high that the Ethernet network is saturated. The spanning tree protocol provides network protection at the LAN level by building a complex hierarchical management structure between all Ethernet devices in the network. A special management protocol is regularly broadcast between each Ethernet device to monitor the

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-2network topology and to detect fault conditions. If one leg or section of the path is interrupted or fails, the spanning tree algorithm allows the Ethernet network to dynamically build another loop-free set of paths. However, depending on the complexity of the Ethernet network, the spanning tree algorithm may be complex to implement, and a significant amount of time may be required for network reconfiguration in the event of a failure on one of the paths.

The trunking protocol provides a means to aggregate multiple links between two Ethernet devices. While normally a loop condition occurs when parallel connections are established between Ethernet devices, trunking eliminates this loop condition by treating the parallel connections as an aggregation group or a single link. Traffic is load shared across the parallel links, and thus trunking makes for a fault tolerant connection. When one or multiple links fail, the aggregation group will converge to a new configuration. However, the bandwidth provided by the failed links is lost. Furthermore, there is no direct expansion of the standard to work with optical transport equipment, such as SONET/SDH equipment. Finally, all links in the aggregate group need to be configured in duplex mode and have the same bandwidth.

SUMMARY

An Ethernet protection system comprises an Ethernet communication device operable to be connected to first and second Ethernet lines forming a parallel connection. The Ethernet communication device is able to select the first Ethernet line and transmit and receive data over the first Ethernet line, and upon detecting a failure in the first Ethernet line, automatically select the second

Ethernet line and transmit and receive data over the second Ethernet line.

An Ethernet protection system for protecting an Ethernet connection defined by first and second Ethernet lines forming a parallel connection comprises a first Ethernet mapper card, a second Ethernet mapper card, and a switching device. The first Ethernet mapper card is operable to be connected to the first Ethernet line, and the second Ethernet mapper card is operable to be

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-3connected to the second Ethernet line. The switching device is connected to the first and second Ethernet mapper cards and a communication line, and is operable to switch a connection to the communication line between the first and second

Ethernet mapper cards.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of an Ethernet protection system protecting an Ethernet connection between two Ethernet devices;

FIGS. 2 and 3 are more detailed block diagrams of the Ethernet protection system;

FIG. 4 is a block diagram of another embodiment of the Ethernet protection system;

FIG. 5 is a flow diagram illustrating line detection failure and switching; and

FIG. 6 is a flow diagram illustrating the activation of revertive and nonrevertive modes based on line quality data.

DETAILED DESCRIPTION

The Ethernet protection system protects an Ethernet line connecting two

Ethernet devices against a line failure or failure of an Ethernet card connected to the line. Fig. 1 provides a block diagram of an Ethernet protection system 10 protecting an Ethernet connection 20 between two Ethernet devices 30 and 40. The first Ethernet device 30 may be a SONET network element capable of AddDrop Multiplexing (ADM), Digital Cross Connection (DCC), or both, and the second Ethernet device 40 may be an Ethernet switch. The network element 30 includes a pair of Ethernet mapper cards 50 and 60, and each Ethernet mapper card 50 and 60 is connected to an Ethernet line 52 and 62, respectively. The Ethernet lines 52 and 62 form a parallel connection between the network element 30 and the switch 40. The network element 30 transmits and receives data over a

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-4SONET/SDH network 70. The SONET/SDH network 70 comprises other SONET/SDH devices and Ethernet devices, such as another network element 80, which also includes an Ethernet mapper card 82 that communicates with another Ethernet switch 90 via an Ethernet line 84.

The SONET/SDH network 70 uses an STS-N frame format to transmit and receive data. The STS-N frame can be divided into two main areas: the transport overhead (TOH) and the synchronous payload envelop (SPE). The SPE can further be divided into the STS path overhead and the payload. The payload contains the data a user desires to transmit and receive over the SONET/SDH network 70. Because the SONET/SDH network 70 is a synchronous network, data is multiplexed into the SPE for transmission. The data can then be transported and switched through the SONET/SDH network 70 without having to be examined or demultiplexed at intermediate nodes.

The network element 30 multiplexes and demultiplexes data transmitted over the SONET/SDH network 70. An exemplary device providing network element functionality is the MCN-7000 Advance Multiple Services Carrier Node from Marconi Communications.

The Ethernet mapper cards 50 and 60 map Ethernet frame data into the SPE data and also receive Ethernet frame data from the SPE data. Thus, the network element 30 and the Ethernet mapper cards 50 and 60 allow users to transport Ethernet traffic across the SONET/SDH network 70. Exemplary Ethernet mapper cards 50 and 60 are the FastE Mapper Ethernet cards from Marconi Communications.

By transporting Ethernet traffic across the SONET/SDH network 70,

Ethernet data may be exchanged between the Ethernet switches 40 and 90.

Instead of a single Ethernet line between the network element 30 and switch 40, however, the Ethernet lines 52 and 62 form a parallel Ethernet connection 20 between the network element 30 and the switch 40. These Ethernet lines 52 and 62 are illustratively a pair of 100BaseT lines. The first line 52 is designated a working line, and the second line 62 is designated a protection line.

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-5Likewise, the first Ethernet mapper card 50 is designated the working Ethernet mapper card, and the second Ethernet mapper card 60 is designated the protection Ethernet mapper card. The combination of the working Ethernet mapper card 50 and the working line 52 is a working channel 54, and the combination of the protection Ethernet mapper card 60 and the protection line 62 is a protection channel 64. During normal operation in which the working channel 54 is active, the working channel 54 is used to transmit and receive Ethernet traffic, and the protection channel 64 is not used to transmit and receive Ethernet traffic. Thus, a forwarding loop is not formed, and the Ethernet switch

40 also does not require to be specially configured to accommodate the Ethernet protection system 10.

The protection channel 64 is not used until a failure is detected in the working channel 54. Should a failure occur in the working line 52, or in the working Ethernet mapper card 50, traffic from the network element 30 to the

Ethernet switch 40 is switched from the working channel 54 to the protection channel 64. When the Ethernet switch 40 receives traffic from the protection line 62, the Ethernet switch 40 will automatically reconfigure to use the protection line 62 instead of the working line 52. Traffic disruption time is proportional to the protection switching time in the network element 30.

The Ethernet protection system 10 may also be configured to be revertive, in which traffic is switched back to the working channel 54 once the working channel 54 is restored.

Figs. 2 and 3 provide more detailed block diagrams of the Ethernet protection system 10. In Fig. 2, the Ethernet mapper cards 50 and 60 are connected to a cross connect switch 100. An exemplary cross connect switch 100 is a digital cross connect device. The cross connect switch 100 is also connected to a communication line 102 over which STS-N data frames are transmitted to and received from line cards in the network element 30. The cross connect switch 100 also comprises switching circuitry 104 that is operable to switch the

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-6connection to the communication line 102 between the working and protection Ethernet mapper cards 50 and 60.

The Ethernet mapper cards 50 and 60 are operable to map Ethernet traffic in the upstream direction (indicated by arrow 110) into SONET payload envelopes. The cross connect switch 100 switches the SONET payloads to line cards operable to transmit data over optical fiber channels. In the downstream direction (indicated by arrow 112), traffic from the line card is sent form the cross connect switch 100 only to the Ethernet mapper connected to the communication line 102. As depicted in Fig. 2, the switching circuitry 104 has selected the working Ethernet mapper card 50 and deselected the protection Ethernet mapper card 60. Thus, Ethernet traffic is sent from the working Ethernet mapper card 50 to the switch 40, as indicated by arrow 122, and traffic is sent from the Ethernet switch 40 to the working Ethernet mapper card 50, as indicated by arrow 120.

The protection Ethernet mapper card 60 does not transmit Ethernet traffic, and receives only the Ethernet broadcast traffic, as indicated by arrow 124.

During normal operation, the working Ethernet mapper card 50 transmits line quality data to the cross connect switch 100. The line quality data indicates the status of the working channel 54. The line quality data value may indicate an OK status, in which case the Ethernet mapper card 50 is functioning normally and data is being transmitted and received over the working line 52. The line quality data value may, on the other hand, indicate a FAIL status, indicating either a failure in the working Ethernet mapper card 50 or a failure in the working line 52.

The line quality data may be transmitted in the TOH portion of the STS-N frame. In the exemplary embodiment of Figs. 1-3, the line quality data is transmitted in a Line Switch Configuration/Line Quality Tag (LSC/LQT) 16-bit data field. The LSC is stored in the most significant byte of the data field and the LSB is stored in the least significant byte. The LQT is inserted by the low level on the Ethernet mapper card 50. The TOH is read by the low level on the cross

0 connect switch 100.

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-7The cross connect switch 100 is responsible for operation of switching between the working Ethernet mapper card 50 and the protection Ethernet mapper card 60. When the cross connect switch 100 receives line quality data indicating that the working Ethernet mapper card 50 or the working line 52 has failed, the cross connect switch 100 switches the connection to the communication line 102 to the protection Ethernet mapper card 60. Thus, the protection Ethernet mapper card 60 is selected and the working Ethernet mapper card 50 is deselected. The cross connect switch 100 then activates the protection Ethernet mapper card 60.

The Ethernet mapper cards 50 and 60 preferably have the same configuration. However, the Ethernet protection system 10 may be configured to automatically configure the protection Ethernet mapper card 60 to the same configuration as the working Ethernet mapper card 50 upon switching the connection to the communication line 102 from the working Ethernet mapper card 50 to the protection Ethernet mapper card 60.

Fig. 3 provides a block diagram of the Ethernet protection system 10 after switching from the working Ethernet mapper card 50 to the protection Ethernet mapper card 60. Ethernet traffic is sent from the protection Ethernet mapper card 60 to the switch 40, as indicated by arrow 122. When the Ethernet switch 40 receives traffic from the protection line 62, the Ethernet switch 40 will automatically reconfigure its MAC to Ethemet-port table to use the protection line 62 instead of the working line 52. The Ethernet switch 40 then sends traffic to the protection Ethernet mapper card 60, as indicated by arrow 120.

Fig. 4 provides a block diagram of another embodiment of the Ethernet protection system 10. In this embodiment, the Ethernet protection system 10 is implemented on a single Ethernet mapper card 70 having two available ports 72 and 74, the first port 72 connected to the working line 52 and the second port 74 connected to the protection line 62. The cross connect switch 100 switches between the working line 52 and the protection line 62 by selecting the corresponding ports 72 and 74 of the Ethernet mapper card 70. While this

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921 embodiment will provide a protection channel 64 for the working channel 54 using only one Ethernet mapper card 70, it will also introduce the possibility of a single point failure in the Ethernet mapper card 70 that may, in turn, prevent

Ethernet traffic from being transmitted over both the working and protection lines

52 and 62.

Switching between the Ethernet mapper cards 50 and 60 may be configured to be revertive or non-revertive. When switching is configured as revertive, the cross connect switch 100 switches the connection to the communication line 102 back to the Ethernet mapper card 50 when the line quality data value indicates a recovery in the working channel 54 from the failure condition. When switching is configured as non-revertive, the cross connect switch 100 will not switch the connection to the communication line 102 back to the working Ethernet mapper card 50.

Fig. 5 provides a flow diagram 400 illustrating dual Ethernet line protection switching that may be configured to be revertive or non-revertive. In step 402, the cross connect switch 100 receives line quality data for the working channel 54. In step 404, the cross connect switch 100 determines whether the line quality data indicates a fault or failure error state in the working channel 54. If there is no fault or failure error state in the working channel 54, the cross connect switch 100 maintains the connection between the communication line 102 and the working Ethernet mapper card 50.

However, if the line quality data indicates a fault or failure in the working channel 54, the cross connect switch 100 switches the connection to the communication line 102 to the protection Ethernet mapper card 60, as shown in step 406. The protection Ethernet mapper card 60 then transmits traffic over the protection channel 64.

In step 408 the cross connect switch 100 determines whether a revertive option has been set. If a revertive option has not been set, line protection processing ends.

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-9If, on the other hand, a revertive option has been set, the cross connect switch 100 enters a wait to restore mode and waits for a time T_w, as shown in step 410. After the wait time T_w has expired, the cross connect switch 100 checks the line quality data for the working channel 54, as shown in step 412. In step 414, the cross connect switch 100 determines if the working channel 54 has recovered. If the working channel 54 has not recovered, steps 410, 412 and 414 are repeated. However, if the working channel 54 has recovered, the cross connect switch 100 switches back to the working channel 54, as shown in step 416. Step 402 is then repeated.

In an alternative embodiment, the Ethernet protection system 10 in the revertive mode enters a wait to restore mode after the recovery of the working channel 54. After the working channel 54 has recovered, the cross connect switch 100 waits for a time T_w before switching the connection to the communication line 102 from the protection Ethernet mapper card 60 back to the working Ethernet mapper card 50.

In the embodiments described above, a default wait to restore time is 5 minutes. The wait to restore time may be adjusted to be more or less than 5 minutes.

Failures in the working channel 54 may also be classified according to failure types, and the Ethernet protection system 10 may be configured to be revertive or non-revertive depending on the failure type indicated by the line quality data. For example, failures may be classified as line failures and card failures, and the Ethernet protection system 10 may be configured to be revertive for a line failure, and non-revertive for a card failure. An exemplary card failure is a power failure, which may be caused by a fault in the Ethernet mapper card 50, by removal of the Ethernet mapper card 50 from the network element 30, or by other events causing loss of power or loss of the Ethernet mapper card 50 processing. An exemplary line failure is an Ethernet UNLINK condition.

Fig. 6 provides a flow diagram 500 illustrating the activation of revertive and non-revertive modes based on exemplary failure types. In step 502, the

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-10working Ethernet mapper card 50 determines whether a fault or failure has been detected in the working channel 54. If no fault or failure has been detected, then in step 504 the line quality data is set to the status OK.

However, if a fault or failure has been detected in the working channel 54, the working Ethernet mapper card 50 determines the fault or failure type. If the fault or failure type is a card failure, the line quality data is set to a card failure value, as shown in step 508. The cross connect switch 100 then switches traffic to the protection channel 64, as shown in step 510, and the switching process then ends.

However, if the fault or failure type is a line failure, such as an Ethernet

UNLINK condition, the mapper card determines whether the failure persists for a specified period of time, as shown in step 512. In the event of a loss of signal (LOS) at the working Ethernet mapper card 50, an UNLINK condition is set. If the UNLINK condition persists for a specified time period, the line quality data may be changed from an OK status to an LOS status. The user may select the duration of the time period. In the embodiment of Figs. 1-3, the time period is defaulted to 200 milliseconds, and may be adjusted from 100 milliseconds to 1 second.

If the failure does not persist, the line quality data is set to the status

OK. If the failure does persist, however, the line quality data is set to a line failure value, as shown in step 514. The Ethernet protection system 10 then switches traffic to the protection channel 64, as shown in step 516, and enters a revertive mode in step 518. The revertive mode is as described with respect to steps 410 - 416 above.

Switching between the working and protection Ethernet mapper cards 50 and 60 may also be implemented manually, such as by a user command from a management interface. For example, if the working Ethernet mapper card 50 is to be temporarily removed or replaced, a user may enter a command that causes the cross connect switch 100 to switch the connection to the communication line

102 from the working Ethernet mapper card 50 to the protection Ethernet mapper

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

-11card 60. Exemplary user commands include Forced Switch To Worker, Forced Switch To Protection, Manual Switch To Worker, Manual Switch To Protection, Clear, and Lock Out Of Protection. Other user commands may also be defined and implemented. The user commands and error states may also be prioritized to override other user commands or override the Ethernet protection system 10 switching conditions resulting from the error states.

The Forced Switch to Worker command causes the cross connect switch 100 to switch to or maintain a connection from the communication line 102 to the working channel 54. This command is prioritized higher than error states in the line quality data and thus traffic remains on the working channel 54 regardless of the error state indicated by the line quality data.

The Forced Switch to Protection command causes the cross connect switch 100 to switch to or maintain a connection from the communication line 102 to the protection channel 64. This command is prioritized higher than error states in the line quality data and thus traffic remains on the protection channel 64 regardless of the error state indicated by the line quality data.

The Manual Switch To Worker command causes the cross connect switch 100 to switch to or maintain a connection from the communication line 102 to the working channel 54 unless a higher priority state occurs. This command is prioritized lower than error states in the line quality data and thus the cross connect switch 100 responds to error states even if this command has been issued.

The Manual Switch To Protection command causes the cross connect switch 100 to switch to or maintain a connection from the communication line

102 to the protection channel 64 unless a higher priority state occurs. This command is prioritized lower than error states in the line quality data and thus the cross connect switch 100 responds to error states even if this command has been issued.

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

- 12The Clear command clears any of the Forced Switch To Worker, Forced Switch To Protection, Manual Switch To Worker, and Manual Switch To Protection commands entered.

The Lock Out Protection Command locks out switching to the protection 5 channel 64. This is the highest priority command and the Ethernet protection system 10 will not switch the communication line 102 from the working channel 54 when this command has been issued.

An exemplary priority scheme for the above-identified commands and error states is provided in Table 1 below. The commands and states are prioritized from 1-10, with 1 being the highest priority and 10 being the lowest priority. The indication field specifies the command or error state. The selected channel field specifies the channel selected when the command or error state is received.

<td> PRIORITY</td><td> INDICATION</td><td> SELECTED CHANNEL</td>

<td> 1</td><td> LOCKOUT OF PROTECTION</td><td> WORKING</td>

<td> 2</td><td> FORCED SWITCH TO PROTECTION</td><td> WORKING</td>

<td> 3</td><td> FORCED SWITCH TO WORKING</td><td> PROTECTION</td>

<td> 4</td><td> SIGNAL FAIL WORKING</td><td> WORKING</td>

<td> 5</td><td> SIGNAL DEGRADE WORKING</td><td> WORKING</td>

<td> 6</td><td> MANUAL SWITCH TO PROTECTION</td><td> WORKING</td>

<td> 7</td><td> MANUAL SWITCH TO WORKING</td><td> PROTECTION</td>

<td> 8</td><td> WAIT TO RESTORE</td><td> PROTECTION</td>

<td> 9</td><td> DO NOT REVERT</td><td> PROTECTION</td>

<td> 10</td><td> NO REQUEST (E.G., WORKING OK AND IN REVERTIVE MODE; WORKING OK; PROTECTION IN NON-REVERTIVE MODE)</td><td> PROTECTION OR WORKING</td>

Table 1 : Command/Error Priority Scheme

According to Table 1, if the Lockout of Protection command has been issued, which is the highest priority command, and a signal failure is detected in the working channel 54, which is a lower priority state, the Ethernet protection system 10 will not switch to the protection channel 54. Conversely, if a Manual

Switch to Working command has been issued, and a signal failure is detected in

CA 02446239 2003-11-03

WO 02/099676

PCT/US02/17921

- 13the working channel 54, which is a higher priority state, the Manual Switch to Working command will be ignored.

The priority scheme, commands and error states of Table 1 are exemplary; other priority schemes, commands and error states may be implemented.

The embodiments described herein are examples of structures, systems or methods having elements corresponding to the elements of the invention recited in the claims. This written description may enable those of ordinary skill in the art to make and use embodiments having alternative elements that likewise correspond to the elements of the invention received in the claims. The intended scope of the invention thus includes other structures, systems or methods that do not differ from the literal language of the claims, and further includes other structures, systems or methods with insubstantial differences from the literal language of the claims.