WO1997001905A1

WO1997001905A1 - Implementation of a fault-tolerant bus in a telecommunications network

Info

Publication number: WO1997001905A1
Application number: PCT/FI1996/000372
Authority: WO
Inventors: Jukka Kainulainen; Mika LINDGRÉN
Original assignee: Nokia Telecommunications Oy
Priority date: 1995-06-28
Filing date: 1996-06-26
Publication date: 1997-01-16
Also published as: FI953205A; FI98582B; EP0840968A1; FI98582C; FI953205A0; AU6226396A

Abstract

The invention relates to implementing a secured bus, particularly a management bus in a telecommunications network comprising a plurality of nodes connected to each other with data connections. In the network, synchronization status messages in accordance with recommendation G.704 are transmitted for indicating the quality level of the signal with respect to synchronization. In accordance with the method, a bus (MB) is implemented in the network for transmitting information outside user traffic, and in securing the bus, a conditional switching point (E) is used in at least one node so that the bus is switched to pass via the switching point, provided that a predetermined criterion related to conditional switching is fulfilled. In order to ensure a simple implementation for securing the bus, the state (QL) of the synchronization status message is used as the criterion so that upon the synchronization status message changing into a state indicating that the signal in question is not to be used for synchronization, the management bus is switched to pass via the conditional switching point.

Description

Implementation of a fault-tolerant bus in a telecom¬ munications network

The invention relates to a method as claimed in the preamble of the attached claim 1 and an arrangement as claimed in the preamble of the attached claim 3 for implementing in a telecommunications network a secured bus for management purposes etc.

A management bus refers to such a data transmission connection which is used for transmitting network management information, that is, information whose purpose is e.g. to collect fault information or to direct functional units of the network into a specific state. In principle, the method of the invention is suited for implementing any kind of bus, although in practice, transmission of management information is by far the most important thinkable implementation of the bus.

The primary object of the present invention is to carry out a secured management connection in a loop network or a so-called mesh network which consists of loops (in a mesh network, there are more than one route between any two nodes), but the invention is generally suited for implementing a secured bus if the form of the bus is suitable for the structure of synchronization hierarchy used in the network. Nodes that are connected to the bus may transmit their messages to the bus, and they will thus spread along the bus to all the nodes connected to it. The bus may not contain a loop, because in that case the transmitted message would remain passing round in the loop (the crossing points of the data connections of a telecommunications network are herein referred to as nodes. Node devices include e.g. branching and multiplexing devices and cross-connection switches) . Signals between the nodes of the network may be e.g. 2048 kbit/s signals in accordance with ITU-T (former CCITT) recommendations G.703/G.704, a frame of said signal containing 32 time-slots (TS0...TS31) and the multiframe containing 16 frames. The management information may be transmitted in the frame structure of such a signal e.g. so that a service channel reserves e.g. three bits of the bits of time-slot TSO from the frame structure. In every other frame, time- slot TSO contains a frame alignment signal, but in every other time-slot bits 4 to 8 are free for domestic use, in which case they may be used for transmitting network management information.

Nowadays, a secured management bus is typically implemented by making use of conditional methods especially implemented for this purpose, in other words, the state of the signal to be transmitted and some specific status flags are monitored in a node of the network. When they meet specific predetermined conditions, a securing connection is switched on. This will be described in more detail in the following.

Figure la illustrates the principle according to which the management bus is typically implemented within a nodal device of the network. The node has a summer 11 to which all the connections of the management bus are applied. These connections may include:

- a local management interface 12 located in the device, to which interface e.g. the network management system may be directly connected. There may be more than just one of these local management interfaces.

- management channels EOC_l...EOC_N (EOC, Embedded Operation Channel or ECC, Embedded Control Channel) transmitted in the different useful signals - a connection to a separate control block 13 of the node, in which block the messages entering the node are processed and from which the received messages are answered.

Messages may arrive from all the directions shown in the figure, and all these directions receive messages (in other words, the traffic from the summer is bidirectional in all the directions shown in the figure) . In a summing point 11, channels coming from the different directions are combined. Data coming from a certain direction is switched (transmitted) further to all the other directions. Only one transmitter at a time may transmit data (in case more transmitters transmit data to the bus simultaneously, the signals will be summed with each other and the data will be corrupted) .

In principle the summer is a very simple device; a single logical AND operation is carried out in it for the management or service channels connected to it. If at least one of the signals is in a "zero" state, the result is zero. This means in practice that the idle mode of the line is one. In other words, if the device is not transmitting anything, it remains in state 1. (The line coming from the control block 13 to the summer is in state 1 when the node has nothing to transmit, whereby it does not have any effect on the other lines) . The traffic is serial from the point of view of the summer 11 and communication blocks of the devices connected to the service channel. At some stage, the data may be in the parallel form, but when applying it to the summer, it must be converted into the serial form.

When securing is carried out in the solution disclosed above, the desired EOC channels must be able to be switched on or off when need be (depending on the states of the signal and the monitored bits) . This may be carried out e.g. in accordance with Figure lb so that there is a separate switch SW for each EOC channel, said switch being controlled on the basis of the states of the signal and of the monitored bits. A signal entering the node is first directed to a multiplexing block MUX/DEMUX, in which the signal of the EOC channel is separated from the rest of the signal and applied to the switch SW. The other time¬ slots of the arriving signal are switched to a cross- connection block XC, from which they are switched further. The control block CTR of the switch receives the information on the state of the signal from the multiplexing block, and it also reads the condition bits from the cross-connection block. On the basis of this information, it switches the securing connection on by commanding the corresponding switch SW on. (For the sake of simplicity, only one signal is shown in the figure, but in the case of more signals, the operation is carried out logically in the same way. On the network level, the above appears in such a way that the network management system has been connected to the management bus in one node via which it has an access to the other nodes of the network along the management bus. This is illustrated in Figure 2, in which the network shown comprises a master node M and eight nodes 1...8, and the network management system NM is connected to node 2. Normal data connections between the nodes are marked with broken lines and the management bus MB with a thicker, solid line. In reality, the management bus thus consists of normal data connections (i.e. it passes within normal data connections), but in the logical sense, however, it is a network that is separate from the normal data transmission connections (as shown in Figure 2) . At certain points of the network, there are conditional connections for the management bus (similar to those in Figure lb) for switching on the securing management connection. In this exemplary case, the conditional connections, which are marked with a reference symbol E, are located in nodes M, 3 and 4, whereby node M is able to switch on a secured connection in the direction of node 1, node 3 to the direction of node 7, and node 4 in the direction of node 6, when need be.

The management bus is constructed in the network as separate switchings. It is determined separately in each node where the management connections (that must be connected to the summer) come from. The determinations are made by the user and they are permanent. The switchings are always valid when the signal containing the management channel is valid. In case it is detected the signal is faulty (it has a fault that causes e.g. an alarm at the far end, or an alarm is received from the far end) , connecting the management channel to the summer is prevented so that the interferences possibly coming from the connection would not interfere with the traffic on the service channel via the summer. In a secured management bus, one or more switching points of the management connection are conditional, that is, a connecting is carried out to the summer only if a specific predetermined condition (e.g. a predetermined state of a monitored bit) is fulfilled. Depending on the device, the condition may be linked to different factors.

The secured management bus implemented in the manner described above is attended by the drawback that the user (operator) has had to create a separate system for securing, in which system a condition bit is transferred. The condition bit remains in a certain state in a normal situation and changes its state at that point of the network at which an unusual situation, such as a fault situation arises.

The object of the present invention is to relieve the drawback mentioned above and achieve a new, simpler solution than heretofore for implementing a management bus or the like. This is achieved with a method of the invention, which is characterized in what will be set forth in the characterizing part of the attached claim 1. The arrangement according to the invention, in turn, is characterized in what will be set forth in the characterizing part of the attached claim 3.

The idea of the invention is to make use of a Synchronization Status Message SSM specified in ITU-T standard G.704 so that the criterion used at the conditional switching point of a secured service channel is linked to the state of this message. In the invention, information intended for the use of synchronization of the network and already existing in the network is thus utilized for a new purpose in the invention.

On account of the solution in accordance with the invention, the criterion information is thus already in the network, and there is thus no need to create any kind of arrangements for transmitting the criterion bit in the network.

In the following, the invention and its advantageous embodiments will be disclosed in more detail with reference to Figures 3a - 6d in the examples in accordance with the associated drawings, in which

Figure la illustrates the principle of implementation of the management bus inside a nodal device of the network,

Figure lb illustrates the principle of implementation of securing the management bus inside a nodal device of the network,

Figure 2 illustrates implementation of the management bus on the network level,

Figures 3a and 3b illustrate the operation of the method of the invention in a ring network,

Figure 4 illustrates those elements of the network node by means of which the method of the invention is implemented,

Figures 5a...5d illustrate the operation of the method of the invention in a so-called mesh network, and

Figures 6a...6d show alternative switching devices for the node shown in Figure 4.

ITU-T standard G.704 specifies the frame structure of a digital transmission system operating at the rate of 2048 kbit/s. In accordance with the recommendation, bits 4 - 8 in every other frame are free (spare bits), and they may be used e.g. for transmitting synchronization status messages mentioned above. Only one of the bits 4 - 8 in the frame may be used for this purpose, and a four-bit synchronization status message thus consists of the selected bit (4-8) in the frames 1, 3, 5, 7 and 9, 11, 13 and 15 of the multiframe. The following table shows the synchronization quality levels (QL) indicated by the bit patterns composed of these selected bits Sanl-San4 (n=4, 5, 6, 7) . In the last column, expressions according to the recommendations have been used.

QL San 1 - Synchronization Quality (QL) San4 Description

0 0000 Quality unknown (Existing Sync. Network) 1 0001 Reserved

2 0010 G.811

3 0011 Reserved

4 0100 G.812 Transit

5 0101 Reserved

6 0110 Reserved

7 0111 Reserved

8 1000 G.812 Local

9 1001 Reserved

10 1010 Reserved

11 1011 Synchronization Equipment Timing Source (SETS)

12 1100 Reserved

13 1101 Reserved

14 1110 Reserved

15 1111 Do not use for Synchronization

As appears from the table, four synchronization levels have been fixed within ITU-T, and two other levels have been provided with a meaning; one of them indicates that the synchronization level is unknown and the other one indicates that the signal is not to be used for synchronization (QL=1111) .

In accordance with an embodiment of the present invention, the latter quality level (15) is used in implementing a secured management bus or the like in such a manner that upon the status message changing from status "XXXX" ("XXXX" being unequal to "1111") into status "1111", the management bus is switched on at the conditional switching point of the network.

Figures 3a and 3b illustrate the operation of the method according to an embodiment of the invention in a ring network comprising a total of five nodes, which are marked with reference symbols N1...N5. The management bus MB has been constructed so that it goes through the entire ring and is cut off beside the master node (Nl) in the direction of node N5. A criterion (reference symbol E) is set to this break point (i.e. to the output port of node Nl in the direction of node N5) . In addition, it is monitored at the normal switching point whether the incoming connection has such a fault due to which the use of the connection for synchronizing must be forbidden. In case such a fault arises, the bus is cut off at this point. By means of this procedure it is ensured that there is not such a point left in the bus from which the messages could return when the conditional switching is switched on. Inside each node, the quality level (QL:XXXX) of the internal clock of the node is marked on top. Below the quality level, there is the synchronization list of the node, the selected timing source being marked in italics on said list. The synchronization method used is of such a type that each node selects as the timing source the signal that contains the synchronization message with the highest quality level. In case several signals have the same quality level, the one highest on the priority list will be selected. Beside each port of the node, the synchronization status message transmitted by the node is marked with reference symbol "QL:XXXX". To the master node Nl and node N3, external timing sources Sl and S3, respectively, are connected. The quality levels (QL=0010 and QL=0100) of the synchronization status messages transmitted by external sources are marked above the sources. Each external source outside the loop synchronization must be provided with a QL value. Figure 3a shows the network in its normal state (no faults) . Master node Nl uses an external timing source Sl, which in this example has been determined as a clock on the level QL=0010. The master node thus transmits this synchronization status message in both directions. The slave nodes are synchronized with a signal received from the port (Pa) on the side of the main direction, the synchronization status message contained by said signal being QL=0010. Consequently, they forward the same quality level (QL=0010) from a port Pb and transmit the quality level QL=1111 ("do not use for synchronization") in the direction from which they receive their timing (in the direction of port Pa) .

Figure 3b shows a situation in which a fault has arisen on the connection between nodes Nl and N2. When node N2 detects this fault, it selects a new timing source. As it is receiving from the other direction (from node N3) the quality level QL=1111, it cannot use this direction for timing either, and it thus changes to the internal timing mode and starts transmitting the quality level QL=1011. The following node (i.e. node N3) receives this quality level from port Pa, whereby it changes the external source S2 in place of its timing source, because the quality level QL=0100 provided by it is better than that received from port Pa and port Pb cannot be used for timing (QL=1111) . Node N3 starts transmitting the quality level QL=0100 in both directions. Node N2 is synchronized with a signal coming from node N3 because the quality level contained by it is better than its internal quality level (QL=1011) , whereby it starts to transmit the quality level QL=1111 in the direction of node N3. Node N4 also accepts the quality level transmitted by node N3 because it is receiving the quality level QL=1111 from port Pb. Node N4 thus transmits the quality level QL=0100 to node N5, which is synchronized in the direction of port Pb, since it provides the quality level QL=0010. Node N5 thus returns the quality level QL=1111 to node Nl and transmits the quality level QL=0010 to node N4. The other nodes in the loop do the same, that is, they transmit the quality level QL=0010 from port Pa and return the quality level QL=1111 to port Pb. It has thus led to the situation shown in Figure 3b. When node Nl detects that the quality level transmitted to it by node N5 changed into level QL=1111 (i.e. node N5 uses the securing direction for its synchronization because there is a fault at some point in the main direction) , it switches the management bus on in the direction of node N5.

Figure 4 illustrates those units of a single network node that are essential for the invention. A node N comprises a plurality of parallel interface units IUl...IUn, by means of which the node is connected to the network (the interface may be e.g. a 2 Mbit/s PCM interface according to the above recommendations, as shown in the figure) , and a control unit CU common to several interface units. Each interface unit comprises at its input a multiplexing/demultiplexing block 81, which is connected via a switch SW to an internal bus IBUS of the node. On this bus, a logical AND operation takes place for signals switched to the bus (in a normal state, pull-up resistors maintain the bus in the state 1 until some of the signals switched to the bus commands the bus to the state 1) . As it was disclosed above, the desired management channels must be able to be switched off and on according to the situation for implementing a secured bus. This may be carried out in accordance with Figure lb so that each management channel has a switch SW of its own, which is controlled from a control block 82 of the interface unit. The EOC management channel is thus separated from the rest of the signal in multiplexing blocks 81, and the control block commands the switch on when the synchronization status message shifts from state "XXXX" into state "1111" ("XXXX" being unequal to "1111") . The value of the synchronization status message is provided to the control block from the multiplexing block 81, which forms the value from the bits that have been selected for this purpose. It is assumed in Figure 4 that a criterion has been set only to the port corresponding to interface unit IUl, and the value of the synchronization status message is thus transmitted to the management bus switching control only in this port (in the other ports said value is transmitted to be used by synchronization only) .

The control unit CU comprises a summer 83 for connecting the management channel coming from a control block 13 of the node and from the network management system NM in a corresponding manner by means of the switch SW to summer 11 (cf. Figure 1) . The network management system is connected by means of a data connection (typically a serial connection, such as V.ll) to a serial port SP at the input of the control unit, said serial port in turn being connected to summer 83.

Figures 5a...5d show the operation of the method of the invention in a network in accordance with Figure 2. In Figure 5a the network is shown in the normal state, in which all the switches of the conditional switching points are in off-position. In the situation shown in Figure 5b, a fault occurs in the master loop A between nodes M and 2. Upon detecting the fault, nodes M and 2 cut the bus off from the faulty connection. In the same way as in the case of Figures 3a and 3b, the nodes of the master loop are synchronized in the opposite direction. When node M detects that the quality level transmitted to it by node 1 has changed into the quality level QL=1111, it switches the management bus on in the direction of node 1. In the case of Figure 5c, a fault also occurs in the slave loop C, whereby nodes (6 and 8) on both sides of the fault cut off the connection. The criterion at the end of the loop in node 3 is triggered when the state of the synchronization message received from node 7 changes from state QL=0010 into state QL=1111. In Figure 5d, a fault occurs in the other subloop (B) , as well. The nodes (5 and 6) at the ends of the faulty connection again cut the bus off from the faulty connection. The criterion at the end of the loop in node 4 is triggered when the state of the synchronization status message received from node 6 changes from state QL=0010 into QL=1111. Figures 6a...6d show various examples for implementing a single interface unit provided with a conditional switching. Figure 6a shows a solution corresponding to that in Figure 4 in which solution the fault and status information are obtained directly from the multiplexing block and in which separate switches SW are used, via which the signal of the management channel is switched to the summer 11 when need be. From the multiplexing block, the signal is switched to the cross-connection block 91 (not shown in Figure 8), from which it is switched further. Figure 6b shows a second alternative in which the entire signal (all the time¬ slots) are applied from the multiplexing block to the cross-connection block 92, from which the EOC channel is connected (when need be) to the summer 11 and the other channels are connected to the other desired points. The QL value is again obtained from the multiplexing block. Figures 6c and 6d show such alternatives in which the control block 82 controlling the switch SW obtains the fault information and the QL value from a separate fault database 93, not directly from the multiplexing block. In other respects, the implementation of Figure 6c thus corresponds to that shown in Figure 6a and the implementation in Figure 6d corresponds to that in Figure 6b. Although the embodiments of the invention have been disclosed above with reference to the examples according to the associated drawings, it is obvious that the invention is not limited thereto, but it may be modified within the scope of the inventive idea set forth above and in the attached claims. Provided that e.g. the recommendations change so that the corresponding information can also be obtained with other bit patterns than QL=1111, such other bit pattern may naturally be used in the same way for implementing a securing of a management bus. What has been disclosed above is only an example of the basic structure of the apparatus, and a person skilled in the art may vary this structure in a plurality of ways without deviating from the idea of the invention.

Claims

Claims :

1. A method for implementing a secured bus, particularly a management bus in a telecommunications network comprising a plurality of nodes connected to each other with data connections, said network being used for transmitting synchronization status messages in accordance with recommendation G.704 for indicating the quality level of a signal with respect to synchronization, according to which method:

- a bus (MB) is implemented in the network for transmitting information outside user traffic, such as information related to network management, and

- in securing the bus, a conditional switching point (E) is used in at least one node so that the bus is switched to pass via said switching point only if a predetermined criterion related to conditional switching is fulfilled, c h a r a c t e r i z e d in that the state of the synchronization status message is used as said criterion so that upon the synchronization status message changing into a state indicating that the signal in question is not to be used for synchronization, the management bus is switched to pass via the conditional switching point.

2. A method as claimed in claim 1, c h a r a c t e r i z e d in that upon the synchronization status message changing from state "XXXX", which is unequal to "1111", into state "1111", said switching is carried out.

3. An arrangement for implementing a bus, particularly a management bus in a telecommunications network comprising a plurality of nodes connected to each other by means of data connections, according to which arrangement synchronization status messages in accordance with recommendation G.704 are transmitted in the network for indicating the quality level of the signal with respect to synchronization, - a bus (MB) is implemented in the network for transmitting information outside professional communication, such as information related to network management, and

- there is a conditional switching point (E) in at least one node of the network so that the bus is switched to pass via said switching point only if a predetermined criterion related to conditional switching is fulfilled, c h a r a c t e r i z e d in that the switch means (82, SW, 92) in said conditional switching point are responsive to said synchronization status message so that upon the synchronization status message changing into a specific predetermined state, the switch means carry out switching of the bus.

4. An arrangement as claimed in claim 1, c h a r a c t e r i z e d in that upon the synchronization status message changing into state "1111" from some other state, the switch means carry out switching of the bus.