SE516103C2 - Method and apparatus in a packet switching network - Google Patents

Abstract

A method and an apparatus for performing protection switching in a packet switched communication network, between at least one active connection and at least one standby connection, wherein information cells are received in at least one ingress port of a node, said method comprising the steps of comparing information found in or in connection to each cell, to state information found in or in connection to the ingress port; accepting or discarding each received cell on the basis of this comparison; when protection switching is to be carried out, changing the state information found in or in connection to the ingress port to perform a protection switching. A connection may belong to several protected domains in a hierarchy of protected domains, or on the same level.

Description

30 516 103 c u o ~ en 2 1: 1 protection switching means that the protected traffic is only sent on one of the connections, ie. either the working connection or the protective connection, from the source.

The selector, at the receiver, must in this case be synchronized with the source to know which connection the track is to be selected from. In this configuration, the protection switching requires action both at the source and at the receiver in the protected domain, which makes this configuration slower than the IH configuration. When protection switching is not performed, the protection connection can be used for other traffic so that the network's resources are utilized better than with the IH configuration. mzn protection switching means that m protection connections can be used by n working connections if a connection ceases to function. If m connecting connection is protected if everyone gets into trouble at the same time.

In ATM, protection switching can be performed at Virtual Path (VP) level or Virtual Channel (VC) level. Multiple connections at the same level (VP or VC) can be grouped in a logical group called Virtual Path group or Virtual Channel group (VPG or VCG). The ability to handle a large number of connections such as one device reduces the processing power required for protection switching.

The protection switch performed by the selector can be performed in your own way. One way is to determine, for each protected connection, whether cells received on the connection should be accepted or not. This can take time if many connections need to be switched at the same time. When a VP or VPG connection is switched, a large number of VC connections must be switched at the same time if a VP or a VPG ceases to exist. The speed of treatment then becomes important. It can also be difficult to keep the operation and maintenance functions active during the standby connection. These actions presuppose that certain actions, e.g. cell transmission, disconnected.

Aim of the invention It is an object of the invention to make the transmission of information in packet transmission networks more reliable. Summary of the Invention Tank This object is achieved according to the invention by a method for performing protection switching in a packet switching network between at least one active connection and at least one standby transmitter connection, wherein information cells are received in at least one input port of a node, which method comprises the steps of: - information contained in or in connection with each cell being compared with state information in or in connection with the input port, - - each received cell being received or discarded on the basis of this comparison, - when protection switching is to be performed the state information is changed in or adjacent to the input port to perform a protective switch.

The object is also achieved according to the invention by a node for receiving information cells in a packet processing network, comprising: - means for extracting information contained in or in connection with each cell, which identifies a domain, or a group of domains, to which cell belongs to, - means for comparing the extracted information with state information found in or adjacent to the entrance gate, - means for determining, on the basis of the comparison, whether the cell should be received or discarded, - means for performing a protection switch by changing the information in the following table.

This method and this device enable rapid protection switching of any number of connections, even when the number of connections is very large.

The information found in or adjacent to the cell preferably identifies a domain and / or a group of domains to which the cell belongs.

The connections that are to belong to the same domain can therefore be chosen freely. for example. regardless of the physical connections on which they are received. 10 15 20 25 30 516 103 The domain information can be added to the cell, e.g. in the entrance gate.

With the method and device according to the invention, performance can be monitored for both the standby connection and the active connection, and whether protective switching is to be performed can be determined from the performance of both connections.

An ATM network is also described comprising at least one ATM node for receiving and switching ATM cells, which ATM node comprises - means for extracting information contained in or adjacent to each cell, which identifies a protected domain , or a group of protected domains, to which the cell belongs, - means for ironing the extracted information with state information contained in or adjacent to the entrance gate, - means for determining, on the basis of the comparison, whether the cell is to be received or discarded, - means for performing a protection switching by changing the information in said table.

Brief description of the drawings The invention will be described in more detail in the following, with reference to the drawings, in which Figure 1 shows the basic principle of protective switching, Figure 2 shows the principle of protective switching according to a first embodiment of the invention, Figure 3 shows the principle of protective switching. second embodiment of the invention, Figure 4a shows a data cell according to a first embodiment of the invention, Figures 4b and 4c show a first and a second preferred embodiment of a table used according to the invention. Detailed description of embodiments Figure 1 shows the basic principle of protection switching. A source 1 sends information to a receiver 3 through a first 5 and a second 7 exchange, e.g. ATM switches. The source 1 and the receiver 3 can e.g. be personal computers, and the information can of course be bidirectional even if only one direction is displayed in the clock.

The information is sent from the source 1 to the first exchange, from which it is sent as two identical fates through a first 9 and a second 11 output port of the first exchange 5 to a first 13 and a second input port of the second exchange 7. A selector 17 determines which of the connections is to be received in the second exchange 7 and passed on to the receiver 3. In the situation shown in Figure 1, the upper connection passing through the first output port 9 and the first input port 13 must be selected. If the quality of the connection deteriorates, the selector 17 can instead be set so that it selects the second information fl which passes through the second output port 11 and the second input port 15.

As is well known in the art, the standby connection can be used to send other information when it is not needed for the protection switching, which enables a more efficient use of the network resources. The configuration will then be slightly different from that shown in Figure 1. Before switching, it must also be ensured that the standby connection has started transmitting the desired information. The first and second output ports 9, 11 can of course be realized as a unit, as can the first and second input ports 13, 15.

Figure 2 shows the principle of protective switching according to a first embodiment of the invention. As in Figure 1, a source 101 sends information to a receiver 103 through a first 105 and a second 107 exchange, e.g. can be ATM switches. The information sent from the source 101 is divided into two identical pieces of information on the way out of the first exchange 106, the two identical pieces of information passing through a first 109 and 109, respectively. a second 11 l output port of the first gear to a first 113 resp. a second 115 input port of the second gear. 10 15 20 25 30 516 103 ø | n o oo The selection of which information fl to be transmitted to the second exchange 107 and on to the receiver 103 is performed in the input port 113, 115, based on information i found in or adjacent to each cell. A monitoring unit 116 in each output port and a monitoring unit 117 in each input port perform monitoring functions. The input forwarding unit ll7 verifies all received cells and accepts only cells received on currently connected connections.

It also adds domain information regarding the protected domain for each cell accepted. This information indicates which domain the cell belongs to. A selector unit 119 determines, based on the domain information added by the input monitoring unit 117, which cells are to be sent through a switch port 121 to the switch 107, and which cells are to be discarded. The selector unit 119 compares this domain information with information contained in a table 123 about which domains are active and which are standby.

One or more of your VPs or VCs may belong to the same domain. All cells on all connections belonging to the same domain can be given the same domain information by the input monitoring unit 117 and will be handled in the same way.

The information used to determine whether a cell should be forwarded or discarded can be found in the cell itself or adjacent to it. In one embodiment, the Virtual Path Indicator / Virtual Channel Indicator (VPI / VCI) is used in each cell. In this case, table 123 must include information for each VP / V C as to whether or not this VP / V C is active. If VP / V C is active, the cell should be forwarded to exchange 107, otherwise it should be discarded.

According to another embodiment, the information is added to the cell, or sent out in connection with the cell, to indicate which protected domain the cell belongs to. A VP or a VP group can be identified in the same way. Table 123 then includes information indicating the codes which mean that a cell should be forwarded or discarded. For 10 15 20 25 30 516 103 nu n v n Q | o o ø ~ ø now 7 to change the way in which the cells belonging to a certain connection are to be handled, only the information in table 123 needs to be changed.

If the active connection and the standby connection are handled by different devices, as shown in Figure 2, in which the two connections are handled by different output and input ports, respectively, the same code can be used for both the active connection and the standby connection. . The monitoring units 116 in the output ports 109, 111 and the monitoring units 117 in the input ports 113, 115 make it possible to monitor both the active connection and the standby connection continuously, in order to determine which has the best quality and whether protective switching is to be performed.

Protective switching can also be performed for other reasons, such as a fault at an underlying level. A VC can e.g. protected switch because the VP or the physical connection to which it belongs is incorrect.

Figure 3 shows a second embodiment of the invention. As before, a source 201 sends information to a receiver 203 through a first 205 and a second 207 exchange. The data fl passes through an output port 211 in which it is, among other things, divided into two data fl fates which are sent on separate connections to an input port 213 in the second exchange 207. The data fl fates are identical. In the input port 213, the cells belonging to the active connection are selected and the cells belonging to the standby connection are discarded.

As in Figure 2, the input port 213 includes a monitoring unit 117 which performs performance monitoring and other functions, and which adds information about the protected domain to which the cell belongs. The selection of cells is performed in a selector unit 219.

The selector unit 219 reads the additional information regarding the connection for each cell and decides on the basis of information contained in a table 223 which cells are to be sent. ~. ; ø. g. 8 on to the gear and which cells are to be discarded. The data fate could of course be divided into two after the exit gate.

Figure 4a shows a cell according to a preferred embodiment of the invention. Typically, the cell has a start field 301 comprising information such as address information and a payload field 303. According to the invention, the cell comprises an additional field 305 which identifies the protected domain to which the cell belongs, or the group of connections. In the cell shown in Figure 4a, the additional field 305 includes the information A. The information is not necessarily linked to the VP / V C to which the cell belongs. An unlimited number of VP / VCs can belong to the same domain and be switched protection together. The solution according to the invention therefore enables rapid protection switching regardless of the number of connections for which protection switching is to be performed.

Figure 4b shows a table such as the table 123 shown in Figure 3 or the table 223 shown in Figure 3 according to a first embodiment of the invention. The table has two columns, a first column 307 for the information found in the additional field 305 in each cell, as shown in Figure 4a, and a second column 309 to show how this information is to be interpreted. In this case, the cells having the information A or D in their additional fields belong to 305 active connections and must therefore be forwarded. Cells with the information B or C in their additional fields 305 belong to standby connections and must therefore be discarded.

In the embodiment shown in Figure 2, two tables 123 are needed, one in each of the input ports 113, 115. These two tables need not contain the same information. For example, if both the active connection and the standby connection are indicated by the letter A, the table belonging to the input port that handles the active connection can indicate that cells containing the letter A in the additional field are to be passed through. The table belonging to the input port that handles the standby connection can show that cells comprising the letter A should be discarded. In order to perform a protection switch, the information in the tables regarding the handling of cells containing the information A in its additional fields must be changed in both tables. 10 15 20 25 30 516 103 .- ~ r I - - ø »- -. n In the embodiment shown in Figure 3, only one table can be used. The active connection and the standby connection must therefore contain different information in the additional field. For example, the letter A can be used for the active connection and the letter B for the standby connection. After protection switching, the letter B would then indicate the active connection. Cells containing the information A would belong to the standby connection and thus be discarded.

Figure 4c shows an embodiment in which the domain A comprises fl your subdomains A1, A2, A3. At certain times, the subdomains belonging to the same domain must be managed together, while at other times it must be possible to manage them individually. In this embodiment, the domain and subdomains are in the same table, with a column 311 indicating the domain resp. the subdomain and a column 313 indicating for each domain or subdomain whether it is active or standby. Two or fl your tables can of course be used for different domain and subdomain levels. In order for a cell belonging to any of the subdomains A1, A2, A3 to be received, the domain A must be active and the relevant subdomain must be active. To change all subdomains to standby mode, only the registration for domain A must be changed to standby.

For cells transferred to unprotected connections, the additional field, and the table, should contain information indicating that these connections should be active.

This information should not be changed.

As mentioned above, in some embodiments the information in the start field 301 can be used, whereby the additional field 305 is not needed. The table shown in Figure 4b will then include VPI / V CI for each protected domain instead of the letters A, B, C and D.

According to one embodiment, the same connection may belong to more than one protected domain.

For example, all connections transmitted on the same physical link may belong to a 10 15 516 103 o o | n a. 10 protected domain while at the same time there is a protected domain e.g. for each trac t class (such as Constant Bit Rate - CBR, Available Bit Rate - ABR, etc.). This would enable e.g. protection switching of high-priority connections, e.g. all real-time connections, while protection switching would not be performed for non-real-time connections on the same physical link. When a cell can belong to more than one protected domain, information indicating all protected domains to which the connection belongs must of course be added to each cell. How the connection is to be handled depending on the state of each domain can be specified as desired. The most applicable solution is probably that a connection can only be active if all protected domains to which it belongs are active.

In the embodiments shown in the diagrams, protective switching has been shown between two ATM switches with direct connection between them. This is only to achieve clarity in the description. Protective switching according to the invention can of course also be performed between switches with one or two other nodes in between, or between two terminals.

For each figure, a one-way connection was discussed. However, the invention can be applied to any number of one-way or two-way connections.

Claims (1)

1. 0 15 20 25 30 516 103 I u o u n a en 11i Patent Claims. A method for performing protection switching in a packet switching network between at least one active connection and at least one standby connection, wherein information cells are received in at least one input port of a node, which method is characterized by the steps of - information contained in or adjacent to each cell is compared with condition information found in or adjacent to the input port, - each received cell is received or discarded based on this iron movement, - protection switching must be performed, the state information is changed in or adjacent to the input port to perform the protection switching. . The method of claim 1, wherein said information found in or adjacent to the cell identifies a domain and / or a group of domains to which the cell belongs. . A method according to claim 2, wherein said information contained in or adjacent to the cell is added to the cell in the input port. . A method according to any one of claims 1-3, wherein said node is an ATM node and the cells are ATM cells. . Method according to one of the preceding claims, wherein the performance of both the standby connection and the active connection is monitored and whether a protective switching is to be performed is determined on the basis of the performance of both connections. . A method according to any one of claims 1-4, wherein a protection switching is performed if a resource at an underlying level ceases to exist. Node for receiving information cells in a packet distribution network, characterized in that it comprises: means for extracting information contained in or adjacent to each cell, which identifies a domain and / or a group of domains means to which the cell belongs, means for comparing the extracted information with state information contained in or adjacent to the input port, means for determining, on the basis of said comparison, whether the cell is to be received or discarded, means for performing a protection switching by changing the information in said table. The node of claim 7, further comprising means in or adjacent to the input port for adding the information contained in or adjacent to the cell. The node of claim 7 or 8, further comprising means for monitoring the performance of an active connection and a standby connection, means for determining whether a protection switch is to be performed between the active connection and the standby connection based on their performance. A node according to any one of claims 7 or 8, further comprising means for determining that a protection switch is to be performed if a resource at an underlying level ceases to function. Node according to one of Claims 8 to 10, characterized in that it is arranged to receive and connect incoming cells. ATM network characterized in that it comprises at least one ATM node for receiving and connecting ATM cells, which node comprises: means for extracting information contained in or in connection with each cell, means for comparing the extracted information with information contained in or adjacent to the input port, means for determining, on the basis of the comparison, whether the cell should be received or discarded, means for performing a protection switching by changing the information in said table.