WO2000018050A1

WO2000018050A1 - Improvements in and relating to synchronisation management of a digital telecommunications network

Info

Publication number: WO2000018050A1
Application number: PCT/GB1999/002804
Authority: WO
Inventors: Robin Douglas Abel
Original assignee: Nokia Networks Oy
Priority date: 1998-09-19
Filing date: 1999-08-25
Publication date: 2000-03-30
Also published as: GB2341749A; GB9820347D0; AU2252900A

Abstract

A method of detecting or anticipating synchronisation faults provided for a digital telecommunications network (1) having a plurality of nodes, each of which obtains a synchronisation signal from a source comprising a selected one of the other nodes. The network is interrogated to obtain data on its topology and, for each node, the identity of the selected respective other node. This information is used to construct a virtual model (12) of the network, and the model is then inspected if any node of the system is to select an alternative synchronisation source. Since the data on the topology of the network and the configuration of all the other nodes have already been obtained, this analysis can be relatively rapidly performed, so that any synchronisation feedback loops in the network can be anticipated before they actually occur.

Description

TITLE: IMPROVEMENTS IN AND RELATING TO SYNCHRONISATION MANAGEMENT OF A DIGITAL TELECOMMUNICATIONS NETWORK

Field of the Invention

This invention relates to digital telecommunications networks, and more specifically to a method and apparatus for anticipating synchronisation faults in such a network.

Background to the Invention

In a. Synchronisation Digital Hierarchy (SDH) telecommunications system, all the nodes of the system have to be synchronised. To that end, the system uses a signal from a very accurate clock (for example a caesium or GPS clock) , known as a G811 source, to provide a synchronisation signal for a large number of nodes not directly connected to the clock.

Each of those nodes uses another node as a source of a synchronisation signal so that the signal from the clock is relayed to each of those nodes via at least one other node. In practice, a number of synchronisation trails are thus set up in the system, each trail being defined by a number of nodes which are arranged in series in relation to the synchronisation signal .

Such a system has to be able to cope with changes, for example modifications to the system or faults, which result in a node in a synchronisation trail no longer being able to receive a synchronisation signal from the previous node in the trail.

To that end, it is known for the nodes to be provided with priority lists of other nodes which act as synchronisation sources so that a node which has lost one synchronisation source can select another. It is also known for the nodes to select synchronisation sources on the basis of Synchronisation Status Messages (SSM) . These messages are encoded in the incoming signal to the node (generally at STM-N level, but it can be encoded into a 2M signal) and essentially describe the quality of the signal with respect to synchronisation. This would generally be one of the following (in descending order of quality) : G.811, G811T, G811L, Internal, DNU (Do Not Use) . The node would select the highest quality incoming source for synchronisation.

Generally, SDH systems employ a mix of these techniques with some nodes using only the priority list approach while others use a combination of SSM and priority lists. In the latter case, which is the most common approach for systems using the optical signal for synchronisation purposes, if there are two incident ^"signals of the same quality selection is based on the priority list.

Ideally, the design of the system is such that any such change in source still results in any node in the system being provided with a synchronisation signal derived directly or indirectly from the accurate clock.

However, it is possible that this would not be achieved in practice if a feedback loop for the synchronisation signal were to occur in the system. The nodes in the feedback loop will no longer be deriving a synchronisation signal from the clock, but will instead try to derive synchronisation signals from each other in a recursive manner. As a result, the node clocks derived from the erroneous synchronisation trail will drift beyond acceptable limits of accuracy, causing a loss of synchronisation of some or all of the clocks of the nodes in the feedback loop. Those nodes will be out of synchronisation not only with the rest of the system, but also with each other, giving rise to a complete loss of synchronisation.

Summary of the Invention

According to a first aspect of the invention, there is provided a method of detecting or anticipating synchronisation faults in the synchronisation network of a digital telecommunications system having a plurality of nodes each of which obtains a synchronisation signal from a source comprising a selected one of the other nodes, the method comprising the steps of:

a) Obtaining data on the topology of the network and, for each node, the identity of the respective selected other node;

b) Constructing a virtual model of the telecommunications network based on that data; and

c) ^'Inspecting said model to determine whether any such faults are prese^'nt in the network, or would be present if the network were to be modified in a specified way.

Once constructed, the virtual model can be consulted whenever a node is to change its selected source. The effect of this change can then be ascertained using the model without, at that stage, having to interrogate the telecommunications network to obtain the necessary information on topology and node configuration. The invention thus provides a relatively rapid means of determining whether a proposed change in the selection of a synchronisation source will result in any synchronisation faults.

Preferably, said inspection of the model comprises identifying a number of nodes which are arranged in series in relation to the synchronisation signal, and therefore define a synchronisation trail along which the synchronisation signal is relayed, the method further comprising the step of determining whether the trail defines a feedback loop for the synchronisation signal.

Because the effects of proposed changes in the selection of a source can be determined relatively quickly using the virtual model, the invention enables such loops to be anticipated in the virtual model before the change of selection occurs in the actual network or at least before any feedback loop in the network is established long enough for synchronisation to be lost .

Preferably, if the inspection of the virtual model reveals any such feedback loop, the configuration of one or more nodes of the telecommunications network is (or can be) altered so as to avoid or remove any such loop in the actual network.

Preferably, an alarm is triggered by the detection of feedback loop in the virtual model .

Preferably, the data on the topology of the network is obtained by interrogating the network. This can be done by sending a trace signal through the network (for example using an RS-TTI technique) so that the information on topology can be obtained automatically.

Preferably, the method includes the step of interrogating the system to determine, for each node, a respective list of other nodes, one of which provides a primary source of synchronisation signal, the other one or more listed nodes providing one or more secondary sources of the synchronisation signal if the primary source is not available.

The inspection of the virtual model is conveniently triggered by a selected primary or secondary synchronisation source for a node becoming unavailable, said inspection identifying any new synchronisation trail which will arise as a result of the selection of another source from the list for that node.

Preferably, the notification of the change of selection is automatically provided by the telecommunications network.

Preferably, said inspection occurs, and any feedback loops are detected in the virtual model , before any change in the selection of a synchronisation source for a node in the telecommunications network.

Preferably, the interrogation of the network to obtain information on the topology and/or said source lists is repeated periodically, and the virtual model updated accordingly.

According to a second aspect of the invention, there is also provided apparatus for anticipating synchronisation errors in a telecommunications network having a plurality of nodes, each of which obtains a synchronisation signal from a source comprising a selected one of the other nodes, the apparatus comprising interrogation means for interrogating the network to obtain information on the topology of the latter and on the identity, for each node, the said respective selected other node, processing means for constructing a virtual model of the network using said information, and inspection means for inspecting the model to identify synchronisation trails, each defined by a series of nodes along which the synchronisation signal is relayed, in the network.

Preferably, the apparatus includes interface means which are operable, if one of the synchronisation trails in the virtual model forms a feedback loop, to alter the configuration of at least one other node defining said trail so as to avoid said loop in the actual network.

The invention also lies in a digital telecommunications network equipped with apparatus according to the second aspect of the invention.

Brief Description of the Drawings

The invention will now be described, by way of example only, with reference to the accompanying drawings in which, Figure 1 is a schematic functional view of a telecommunications system fitted with apparatus in accordance with the invention;

Figure 2 is a block diagram which shows the connection between the system and the apparatus .

Figure 3 is a diagrammatic view of a simplified digital communications network, showing interconnections between various nodes of the network; and

Figure 4 is a diagram which corresponds to Figure 2 and shows possible synchronisation trails within the simplified network.

Detailed ^"Description

With reference to Figures 1 and 2 , an SDH digital communications network 1 has a network management system which is connected to the synchronisation management tool, generally referenced 2 , and converses with a number of nodes on the telecommunications network via the Q3 stack for the nodes. The management system is situated at a management station 3 (Figure 2 ) which is also provided with a Man-Machine Interface via which an operator can alter the configuration of the network 1 and be presented with information by the synchronisation management tool 2.

The management station 3 is connected to the telecommunications system via a head end node 5 in the latter. The head end node 5 may be a dedicated node or a conventional node which has been assigned the responsibility of providing the connection to the station 3. The station 3 is responsible for general management functionality (of the telecommunications network) as well as for the management of the synchronisation network. All management communications take place via the Data Communications Network (DCN) in the telecommunications network.

The station 3 is arranged to store various different types of information including control software which is generally indicated by reference numeral 7 in Figure 2, and which provides various functional software tools.

Thus, the system includes a synchronisation management mediation tool 4 which provides a functional interface between the management system 2 and the network management system for the physical network 1. The mediation tool 4 is controlled by a configuration management tool 6 and one of its functions is to act as a means for interrogating the network 1 to determine the topology of the latter. It does this by means of Regenerator Section Trail Trail Identification (RS-TTI) techniques to determine how the nodes of the system are interconnected. The information obtained by this activity is stored in a transmission network data base 8. One example of such a technique will now be outlined.

The signals carried by the SDH network have several constituent overhead sections, one of which is termed the Regenerator Section (RS) overhead. This overhead is of particular importance to the network discovery function in the synchronisation management system as every STM-N node successively accesses and processes the information; this is not necessarily true of other section overheads in the SDH signal. Part of the RS overhead allows the operator to introduce a "signature", such as a text string, into the signal; this is the Regenerator Section Trail Trace Identifier - (RS-TTI) . As part of the general management functionality within SDH, it is possible to set (and determine) the RS-TTI transmitted from each node, and to determine the RS-TTI received by each node . By monitoring the transmitted and received RS-TTI at every node in the network it is possible to construct a model of all physical interconnects within the SDH network.

The synchronisation management mediation tool is also operable to interrogate each node of the system in turn so as to determine its respective synchronisation source list, i.e. its prioritised list of other nodes (which could be represented by the identity of the relevant input, to the node in question) which can act as synchronisation sources for the node. The source lists are then stored on a node synchronisation source list data base 10.

The configuration management tool 6 causes the synchronisation management mediation tool 4 to perform both types of interrogation at regular intervals, for example 2 to 3 days. In addition, the configuration management tool 6 is arranged to cause the synchronisation management mediation tool 4 to perform both sorts of interrogations on additional occasions in response to appropriate commands from the Man-Machine Interface. Thus, for example, an operator might wish to initiate an update if he/she suspects that a node or another feature of the telecommunications network has been altered. For example, if it is known that a particular node has required on-site maintenance for some reason, it would be sensible for the operator to re-confirm the synchronisation configuration of the node subsequent to this. Likewise, if it is thought or known that the physical network has required maintenance of some description (e.g. after a fibre break has occurred) it is again reasonable to presume that the network topology information held in the synchronisation management systems should be checked for accuracy.

It is believed that, if the telecommunications system has around 100 nodes, the interrogations will both take approximately 2 to 3 hours to complete.

The configuration management tool 6 uses the information from the data bases 8 and 10 to construct a virtual model 12 of the network 1. That model not only indicates the various interconnections between the nodes of the network 1, but also indicates possible synchronisation trails, i.e. trails for synchronisation signals, in the network 1. Data describing those trails is stored in a synchronisation trail data base 14.

In the course of the design of the telecommunications network 1, a synchronisation plan will have been conceived in which primary, or first choice, synchronisation trails are implemented by configuring each node to accept one input as a synchronisation source. The network planner will generally also build in a synchronisation protection scheme by defining a secondary source (or sources) which the node will use in the event of the primary source becoming unavailable, for example as a result of the node providing the primary source failing or the failure of the interconnecting link between those nodes. These facilities are implemented in each node by means of its respective source priority list. The priority list of each node is generally manually configured by the network operator, who will endeavour to ensure that no feedback loops occur at least in the primary synchronisation trails. However, as indicated in the introduction of this specification, the system may be designated in such a way that each node uses SSMs to determine which other node is to act on its synchronisation source, or to adopt a hybrid of these two approaches. The management system can be configured to obtain priority lists from any of these types of network.

In any case, when the network is initially set up, no synchronisation feedback loops should be present, and the management system 2 should not interfere with the network other than to perform the interrogations described above.

However, if one of the nodes of the system loses its primary synchronisation source, the network management system notifies the synchronisation management mediation tool 4 which triggers a fault management tool 16 and the configuration management tool 6. Firstly, the model 12 will be updated by the tool 6 to take into account the change in synchronisation source for the node, and hence the alteration of at least one of the synchronisation trails. The updated set of trails is then stored in the data base 14, which is then consulted by the fault management tool 16 to determine whether any of the trails forms a feedback loop.

If any such loop is detected anywhere in the system, the fault management tool 16 sends a suitable warning to the configuration management tool 6, which in turn causes the MMI to notify the operator of the potential problem. In a modified version of the synchronisation management system, the tool 6 also instructs the network management system of the network 1 (via the tool 4) to place the affected node(s) in a hold-over mode^".

The detection of synchronisation loops is achieved algorithmically by analysing the model 12, without the need at that time to undertake the relatively lengthy process of interrogating the network 1. Thus, the management system 2 rapidly ascertains whether a proposed change in source for one or more nodes in the network 1 would give rise to a feedback loop, and can prevent this happening before the loop occurs in the actual network 1.

The fault management tool 16 also works with the configuration management tool 6 to ensure that discrepancies in the network either do not occur or do not affect network functionality. An example of such a discrepancy would be a discrepancy between the source list of a node and the topology of the network.

The exact requirements and modus operandi of the procedure of preparing the priority list initially are defined in the ETSI document "ETSI DTR/TM-03080 Synchronisation Network Engineering" [TM-03080] . Reference is also made to the ETSI DOCUMENT ETS 300 462-2. "Transmission & Multiplexing [TM] : Generic Requirements for synchronisation networks; Part 2: Synchronisation network architecture". Once warned of the risk of a feedback loop, the operator can appropriately alter the configuration of the node(s) in question so that the loop is avoided. Instructions to make this alteration are supplied to the network through the management tool 2 which also updates the virtual model 12 accordingly. The virtual model 12 will subsequently be inspected if another node has to change its synchronisation source, or if the interrogation of the network indicates any change in the topology or source lists.

The management system 2 can also be used in the planning of the configuration of the network by editing the virtual model 12 and ^'then effecting this change on the real network; this is accomplished by the mediation tool 4 re-configuring the appropriate node(s) according to changes made to the virtual model .

Figure 2 represents a simplified version of an SDH telecommunications network 1 which may be connected to the management system 2. The simplified system consists of thirteen SDH modes (A to L plus PI) connected as shown. All have essentially the same functionality and configuration, except for Node PI which is directly connected to a PRC clock (e.g. a GPS receiver or caesium oscillator) of quality G.811 and acts as the master clock for the network. Table 1 shows interconnection data for the network of Figure 2 which may be obtained using the method described above. It is a relatively straightforward task to interpret this information to obtain a network topology map. An advantage of the synchronisation management system 2 is that this information can be obtained and maintained automatically (using the RS-TTI technique) ; there is no reliance on the user or operator to input the information correctly. Table 1

Knowledge of the synchronisation plan and strategy can be determined by examining the priority lists of each node. Table 2 illustrates priority lists which may be defined for each node in the network described above .

Table 2

By correlating this information we can obtain the following information about the primary Synchronisation Trails within the network :

Primary Synchronisation Trail 1

PI → A → B → C

Primary Synchronisation Trail 2

PI → K → E → D

Primary Synchronisation Trail 3

Pl → L → J → I → H

Primary Synchronisation Trail 4

Pl → K → E → F →G

As we also have knowledge of the secondary synchronisation source selections configured on each node, we can now map the topological and synchronisation information to give a virtual model as shown in Figure 3. The resulting logical network (s) of synchronisation trails is known as a Synchronisation Network. The synchronisation network management system manages the virtual network 12; the mediation function 4 then uses the NMS to manage the physical network.

Use of this "virtual" Synchronisation Network is important if responses to changes are to be processed and dealt with acceptable delay. It is estimated that in a network having about 100 nodes,, about 4 hours is required to obtain and process the above data - this is a prohibitive delay for effective management of the synchronisation network. By using the method in accordance with the invention, the reaction time of the tool to changes and/or errors in the synchronisation network is drastically reduced.

Consider now the case where a primary synchronisation source for a node becomes unavailable. If the primary synchronisation source for Node J fails (either because Node L fails of the link between Node L → Node J fails, for example) then Node J will choose its secondary source (Node I) for synchronisation and inform the synchronisation management system 2 accordingly. The system 2 will receive this notification and consults the virtual model 12 which indicates that Node J will now take the input from Node I for synchronisation. It can also be seen from Figure 2 that this effect will result in a new synchronisation trail of Pl → K → E → F → G→ H → I → J (the mechanism of how this re-arrangement should take place is defined in the ETSI document "ETSI DTR/TM-03080 Synchronisation Network Engineering") . This effect can be determined (and displayed to the operator via the configuration management tool 6 extremely quickly, as the management tool 6 is referencing the virtual network only. A "sanity-check" can subsequently be performed, as Nodes I and H will also inform the management tool of a change in synchronisation status.

The scenario presented above presents no real threat to the network functionality, as the synchronisation network has been correctly planned. In this instance, the invention provides "nice to have" features. However, if the synchronisation network has been incorrectly planned, or the nodes' synchronisation configuration has been changed after the network has been initially configured, the invention will prove extremely useful. Consider now the case where node E of Figure 2 if re-configured to accept Node I as its secondary synchronisation source after initial installation; it can be seen that there is a potential synchronisation loop (Node E → Node F → Node G → Node H → Node I → Node E) if the primary synchronisation source for Node E fails. When the synchronisation management system 2 performs its periodic (or operator or source failure instigated) maintenance and inspection of the virtual synchronisation model, this circular dependency will be detected and the operated warned. Furthermore, the fault and configuration management tools can liaise with the mediation tool 4 to prevent Node E accepting a synchronisation input from Node I if a loop situation occurs during network operation. If the situation arises that is a closed synchronisation trail from Node E to Node I, the tools will prevent Node E from using Node I as a synchronisation source. It should be noted that there is a delay in this procedure as it will require the synchronisation configuration management tool to assess the relevant node's synchronisation source configuration data, and as such it is not the preferred operational mode.

It is also possible to re-configure the physical synchronisation network using the system 2. In this case, the virtual model 12 is re-configured (via a configuration edit tool) by the user, and the effect would be realised by downloading the information to each node via the mediation tool. Since the system 2 can detect synchronisation loop potential, the operator could be warned before download takes place. For example, the virtual synchronisation model 12 could be re-configured and the change implemented on the physical network by the mediation function causing the synchronisation source priority lists for each node (i.e. Table 2) to be updated accordingly. In summary, the described embodiment of the invention provides a management system in which a "Virtual synchronisation network" is created by obtaining data pertaining to the physical topology of the real network, and the synchronisation configuration of each node therein.

Virtual management is carried out by a fault management and a configuration management tool (6, 16) which act upon the virtual synchronisation network. Changes and events in the physical network are conveyed to the virtual network by a mediation function between the Network Management System for the physical network, and the management system for the virtual synchronisation network.

Maintenance of the synchronisation network can be automatic if required.

This allows almost immediate (and in some case prior) response to synchronisation errors or failures in the network as fault and configuration management actions are performed abstractly.

A response can be triggered automatically by event notification from the physical network (via the mediation tool 4) , by operator input or by the Synchronisation Management System itself detecting actual or potential errors.

Potential synchronisation loop conditions can be identified before they occur. This can be either in response to the detection of a change in the physical network or during reconfiguration of the synchronisation network by the operator.

In addition, "off-line" reconfiguration of the system can be achieved by modifying the virtual synchronisation network at the management stations, and then mapping this change to the real network. The synchronisation source configuration information of appropriate modes (e.g. priority lists and/or source selection methods) , modified via a Graphical User Interface (GUI) , and the new configuration information is then downloaded to the affected nodes. The synchronisation network configuration can be easily altered in this manner; a "no sync loop" error check can be implemented prior to download.

Claims

1. A method of detecting or anticipating synchronisation faults in the synchronisation network of a digital telecommunications system having a plurality of nodes, each of which obtains a synchronisation signal from a source comprising a selected one of the other nodes, the method comprising the steps of:

a) obtaining data on the topology of the network and, for each node, the identity of the selected respective other node ;

b) constructing a virtual model of the telecommunications system, based on that data; and

c) inspecting said model to determine any synchronisation faults or potential faults in the telecommunications system.

2. A method according to claim 1, in which said inspection of the model comprises identifying a number of nodes arranged in series in relation to the synchronisation signal which is therefore relayed along a synchronisation trail defined by the nodes, and determining whether said trail defines a feedback loop for the synchronisation signal .

3. A method according to claim 2 , in which if a feedback loop is detected in the virtual model, allowing the configuration of at least one of the nodes in the series in the telecommunications system so as to prevent said feedback loop occurring in the system.

4. A method according to claim 2 or 3 , in which an alarm is triggered if a feedback loop is detected.

5. A method according to any of the preceding claims, in which the data on the topology of the system is obtained by interrogating the system.

6. A method according to any of the preceding claims, in which, for each node, the system obtains a respective list of other nodes, one of which provides the primary source of a synchronisation signal for the node, the other one or more listed nodes providing one or more secondary sources of a synchronisation signal for the node if the primary source is not available .

7. ^* A method according to claim 6, in which the telecommunications system is interrogated to obtain said lists for use in the virtual model .

8. A method according to claim 6 or claim 7, in which the inspection of the virtual model is triggered by a selected primary or secondary synchronisation source for a node becoming unavailable, said inspection identifying any new synchronisation trail which will arise as a result of the selection of another source from the list for that node.

9. A method according to claim 8, in which said inspection occurs and any feedback loop for synchronisation signals is detected in the virtual model before any change in the selection synchronisation in the telecommunications system occurs .

10. A method according to claim 5 or claim 7, in which said interrogation of the system to obtain said information on topology and/or said lists.

11. Apparatus for anticipating synchronisation errors in a telecommunications network having a plurality of nodes, each of which obtains a synchronisation signal from source comprising a selected one of the other nodes, the apparatus comprising interrogation means for interrogating the network to obtain information on the topology of the latter and on the identity, for each node of said respective selected other node, modelling means for constructing a virtual model of the network using said information, and inspection means, for inspecting the model to identify series of nodes which define trails for synchronisation signals in the network.

12. Apparatus according to claim 11, in which apparatus includes interface means operable, if one of the synchronisation trails in the virtual model forms a feedback loop, to alter the configuration of at least one of the nodes defining said trail so as to avoid said loop in the actual network.

13. A digital telecommunications network having apparatus according to claim 11 or claim 12.