WO2010075898A1

WO2010075898A1 - Alarm propagation

Info

Publication number: WO2010075898A1
Application number: PCT/EP2008/068374
Authority: WO
Inventors: Karl-Friedrich Rothermel
Original assignee: Nokia Siemens Networks Oy
Priority date: 2008-12-30
Filing date: 2008-12-30
Publication date: 2010-07-08

Abstract

The present invention relates to alarm propagation and, in particular, to providing a more flexible alarm propagation mechanism from network elements (402) to the network operator. The methods and apparatus described herein determine for each of a set of network element alarm statuses a portion of network elements (402) in a map (401) of network elements having each network element alarm status. Based on the portion of network elements (402) having each network element alarm status an alarm state of the map (401) may be determined for each network element alarm status and an overall alarm state of the map (401) may then be determined as the maximum severity alarm state of the alarm states determined for the map (402) for each network element alarm status.

Description

Alarm Propagation

The present invention relates to alarm propagation and, in particular, to providing a more flexible alarm propagation mechanism from network elements to the network operator.

Network Operator' s networks are typically managed using a combination of Network Management Systems (NMS) and Element Management Systems (EMS) . Typically one of the functions of the NMS and the EMS is to provide an indication of the alarm state of each of the network elements and also of the network in general to the network operator.

In terms of a typical telecommunication network the alarm state is defined in the standard ITU-T Rec. X.731 (see h11p : / /www .itu.int/rec /T-REC-X.731 /en ) . The alarm state of a given network element corresponds to the alarm with the high- est severity which is affecting the given network element. In ITU-T Rec. X.733 (see http : / /www .1 tu . i nt / rec/T-REC- X .733 /en ) the alarm state severities are defined as being CRITICAL, MAJOR, MINOR, WARNING and NORMAL.

In order to provide the network operator with an indication of the alarm state of the network elements the NMS and/or the EMS receive alarm events or notifications from the network elements and typically show the alarm state of the network elements to the network operator using a graphical map repre- sentation .

The alarm state shown in the graphical map representation to the network operator is the maximum alarm state of the entities contained within the particular map view, for example, network elements, any sub maps, and so on. If one network element is in alarm state CRITICAL, then the alarm state of the whole map is also shown as CRITICAL.

Accordingly, in the conventional systems the highest severity alarm affecting a network element is propagated from the network element to the NMS and/or the EMS which is then displayed to the Network Operator. However, this conventional approach of propagating and displaying the highest severity alarm state of each of the network elements has several disadvantages and problems. In particular, the conventional alarm propagation mechanism from the network elements to the network operator does not take into account the redundancy mechanisms related with new network architectures.

Network architectures are becoming more complex and the need for redundancy and resilience is a major requirement for network operators to ensure that network service availability is maintained. Typically, a high degree of service availability is achieved by providing resilience through a distributed and replicated network architecture rather than by using expensive proprietary and fault tolerant hardware platforms.

Accordingly, the conventional approach described above for propagating the highest severity alarm from each network element to the network operator does not take into account the new distributed and replicated network architectures. In other words, if the network architecture employs a redundancy scheme which, for example, includes network elements being provided in triplicate typically based at different sites for implementing a particular service or functionality, then if one of those network elements is CRITICAL the remaining two network elements would still be able to provide the service/functionality. However, in the conventional systems the network operator would only see a CRITICAL alarm state and therefore send service engineers immediately to fix the problem when, in fact, corrective measures are not required urgently due to the redundancy scheme implemented and the net- work architecture.

Thus, from a service availability point of view the alarm state is not CRITICAL as the service is still available due to the redundancy scheme implemented in this example. There- fore, the network operator does not need to perform urgent corrective measures as the error can be resolved and corrected at a later date.

In other words, the present invention describes a flexible alarm propagation mechanism which reflects the alarm state of the service availability and functionality rather than only the alarm state of the network element. Moreover, the network operator can adapt the alarm propagation mechanism by defining various parameters which reflect the network redun- dancy schemes implemented.

The following example illustrates the disadvantages and problems with the convention systems and the conventional alarm propagation mechanisms. This example is described with ref- erence to Figure 1 which shows an example of a logical level and a physical level implementation for a central database server, and with reference to Figure 2 which shows an example arrangement of the physical layer shown in Figure 1.

Figure 1 illustrates a logical level 110 which represents the services and functionality provided by the network and the corresponding physical level 111 which represents the actual hardware that implements the logical level. In this example, the logical level includes a Central Database Server 101 which stores varying pieces of information that are used by various application logic 102. Examples of such application logic 102 are a Home Location Register (HLR) or a Home Subscriber Server (HSS) . The logical level in this example also includes the functionality and services for a Provisioning System 103, a business Support System 104 and the EMS/NMS 105.

On a physical level the Central Database Server 101 is implemented using a set of Directory Servers 106 and the Application Logic is implemented using Application Logic Front Ends 107, e.g. HLR Front End.

A network operator can implement any type of redundancy scheme in order to protect the service and functionality availability. In this example, a first redundancy scheme 108 is assigned to the sets of Directory Servers 106 and a second redundancy scheme 109 is assigned to the Application Logic Front Ends 107.

The redundancy scheme 108 assigned to the Directory Servers (DS) 106 in this example is shown in more detail in Figure 2.

As shown in Figure 2, the DSs 207a, b, c, 208a, b, c, 209a, b, c are arranged in arranged in Directory Server Agent (DSA) groups 204, 205, 206. Each of the DSA groups 204, 205, 206 contains three DSs located at three different sites 201, 202, 203 and each DSA 204, 205, 206 implements particular services and functionality in the network.

Considering the first DSA group 204, it contains a first DS 207a at a first site 201, a second DS 207b at a second site 202 and a third DS 207c at a third site 203. If one DS, for example 207a, is out of service (e.g. due to an error or during upgrade of a DS) , the remaining two DSs 207b at site 202 and 207c at site 203 are still available to implement the required functionality and services.

Thus, if the DS 207a has an alarm state of CRITICAL the remaining two DSs 207b and 207c can still implement the required functionality and services. In other words, even though one DS may be CRITICAL the service or functionality provided by the DS is not CRITICAL due to the redundancy scheme implemented in the network architecture.

However, in the conventional systems and the alarm propagation mechanism implemented in the conventional systems the alarm state of the map containing entities is equivalent to the highest severity alarm in that map. Accordingly, if one network element is CRITICAL then the graphical map representation provided to the network operator will be CRITICAL even though the services will be available due to the redundancy scheme implemented.

Thus, it is not necessary to send an emergency repair engineer out to the first site 201 to repair the DS 207a immediately which would incur significant costs but the repair could be provisioned for a time when the cost of repairs is minimised .

Therefore, there is a need for a more flexible alarm propagation mechanism which takes into account the redundancy scheme and network architecture implemented by a network operator.

According to a first aspect of the present invention there is provided a method comprising determining for each of a set of network element alarm statuses a portion of network elements in a map of network elements having the network element alarm status; determining an alarm state of the map for each of the network element alarm statuses based on the portion of network elements having the network element alarm status; and determining an overall alarm state of the map as the maximum severity alarm state of the alarm states determined for the map for each network element alarm status .

Accordingly, for a map of network elements the present inven- tion determines the portion of network elements having each network element alarm status and determines an alarm state of the map for each network element alarm status. Once an alarm state for the map has been determined for each of the network element alarm statuses then an overall alarm state of the map can be determined as the maximum severity of all of the alarm states that have been determined for the map. Thus, the present invention may advantageously enable an overall alarm state of the map to be determined which indicates or represents the state of the map of network elements rather than the alarm status of each individual network element. The map of network elements may provide particular services and functionality and thus it is advantageous to determine the alarm state of those services and functionality rather than simply determine the alarm status of each individual network ele- ment.

The step of determining the alarm state of the map for each of the network element alarm statuses may further comprise determining if the portion of network elements in the map with the network element alarm status is one and setting the alarm state for the map for the network element alarm status equal to the network element alarm status . The step of determining the alarm state of the map for each of the network element alarm statuses may further comprise determining if the portion of network elements in the map with the network element alarm status is zero and setting the alarm state for the map for the network element alarm status equal to NORMAL.

The step of determining the alarm state of the map for each of the network element alarm statuses may further comprise identifying two predetermined boundaries between which the portion of network elements in the map with the network element alarm status is located; and setting the alarm state for the map for the network element alarm status equal to the network element alarm status being reduced by the numerically ordered number of an upper boundary of the two boundaries between which the portion of network elements in the map with the network element alarm status is located.

Accordingly, depending on the determined portion of the network elements in a map having each of the network element alarm statuses the effect or impact of the portion of the network elements having each network element alarm status on the services and functionality provided by the map can be determined. If the portion of network elements having a particular network element alarm status is between two predetermined or predefined boundaries then the impact of that portion of network elements having the network element alarm status may be determined as the network element alarm status being considered reduced by the numerically ordered number of the upper boundary. For example, if the portion of network elements having a particular network element alarm status lies between the third and fourth predetermined boundaries then the alarm state of the map may be determined as the network element alarm status being reduced by three as the upper boundary is the third boundary and therefore the numerically ordered number of the upper boundary is three. The method may further comprise setting a series of boundaries between 1 and 0 wherein the boundaries define a network architecture and redundancy scheme. The series of boundaries may also be numerically ordered starting with bO representing 1 and ending with boundary bN representing 0.

Accordingly, by setting the series of boundaries the network operator can define any network architecture and any redundancy scheme that the network operator may implement in the map of network elements. Therefore, the method may also take into account the redundancy scheme and network architecture when determining the impact of a network element alarm status for a given portion of the map on the services and functionality provided by the map. The boundaries may be a value be- tween 0 and 1 which defines the points at which the portions of network elements with a given network element alarm status has a greater or lower effect or impact on the services and functionality provided by the map.

The portion is determined as the ratio of the network elements having the network element alarm status to the total number of network elements in the map. Thus, the portion of network elements may be calculated such that it falls between 1 and 0 where the value one represents all network elements in the map and zero represents none of the network elements in the map.

According to a second aspect of the present invention there is provided a server configured to determine for each of a set of network element alarm statuses a portion of network elements in a map of network elements having the network element alarm status; determine an alarm state of the map for each of the network element alarm statuses based on the portion of network elements having the network element alarm status; and determine an overall alarm state of the map as the maximum severity alarm state determined for the map for each network element alarm status.

According to a third aspect of the present invention there is provided a computer program product comprising computer readable executable code for determining for each of a set of network element alarm statuses a portion of network elements in a map of network elements having the network element alarm status; determining an alarm state of the map for each of the network element alarm statuses based on the portion of network elements having the network element alarm status; and determining an overall alarm state of the map as the maximum severity alarm state determined for the map for each network element alarm status.

The method described herein may be implemented using processors, memory and so forth as would be appreciated by a person skilled in the art. Furthermore, the present invention pro- vides a transformation of individual network element alarm statuses into an overall alarm state of a map that preferably comprises at least two network elements.

Preferred embodiments of the present invention will now be described, by way of example only, and with reference to the accompanying drawings, in which:

Figure 1 shows an example of a logical level and a physical level implementation for a central database server.

Figure 2 shows an example arrangement of the physical layer shown in Figure 1. Figure 3 shows a typical telecommunications network arrangement .

Figure 4 shows an example map arrangement of network ele- ments.

A typical telecommunication network 301 is shown in Figure 3. A typical telecommunication network includes many network elements 302 which can be arranged in any configuration re- quired by the network operator. The typical telecommunication network also includes an EMS 303 and an NMS 304 which are used to manage the network elements and the telecommunication network in general.

The EMS 303 and the NMS 304 have several functions in order to manage the network elements (NE) 302 and the telecommunication network 301 in general. These functions include grouping NEs 302 into maps and sub-maps where the maps and sub-maps typically reflect regional or functional aspects of the telecommunication network as well as redundancy mechanisms implemented by the network operator. The EMS 303 and the NMS 304 also typically include event managers that collate and present the events and alarms that occur in the NEs to the network operator in a graphical representation of the maps and sub-maps of the network.

The flexible and configurable alarm propagation mechanism of the present invention will now be described with reference to Figure 4. The embodiments will also be described in relation to the ITU-T Rec. X.731/733 NE alarm statuses for a telecommunication network. However, a person skilled in the art will appreciate that the embodiments may be used in any type of network with the alarm statuses defined for that type of network . In order to determine the alarm state of a map that comprises any number of NEs arranged in any network architecture and redundancy scheme the network operator will define the impact that any number of NEs with a particular alarm status will have on the services and functionality provided by the map. This is achieved by setting a series of boundaries reflect the impact that any number of NEs having a particular NE alarm status based on the network architecture and redundancy scheme implemented in the network.

The boundaries that can be set by a network operator lie in the range between 1 and 0, where 1 represents all NEs in the map and 0 represents none of the NEs in the map. The portion of NEs with a particular NE alarm status is determined as a ratio of the NEs affected by a particular NE alarm status to the total number of NEs in the map.

Accordingly, the network operator may set any number of boundaries between the range of 1 and 0 that reflect the network architecture and redundancy scheme implemented by the network operator. The portion of the NEs in the map with a given alarm status can then be determined and the impact of the alarm status of the identified portion of NEs in the map on the services or functionality provided by the map can be determined and reflected in an overall alarm state of the map that may then be provided to the network operator.

The boundaries set by the network operator are ordered be- tween boundary zero which is set at 1 (which represents all of the NEs in the map) through to boundary N which is set at 0 (which represents none of the NEs in the map) . Therefore, if, for example, the network operator predefines two further intermediate boundaries in order to reflect the network ar- chitecture and redundancy scheme implemented then the set of boundaries will include boundary zero being 1, boundary one being the higher of the two boundaries set by the network operator, boundary two being the next boundary of the two boundaries set by the network operator and boundary three being 0.

Each severity of NE alarm status from WARNING to CRITICAL may then be considered in turn and the portion of NEs in the map with each alarm status is determined. Then for each alarm status the portion of NEs with that alarm status is used to determine the alarm state of the map for that particular NE alarm status. Once the alarm state of the map for each of the NE alarm statuses from WARNING to CRITICAL has been de- termined then the overall alarm state of the map may be determined as the maximum severity alarm state of all of the determined alarm states for the map.

When considering each NE alarm status in turn, if the deter- mined portion of the NEs with a particular NE alarm status is 1 then all of the NEs in the map are affected by the particular alarm status. Accordingly, the alarm state of the map for that particular NE alarm status will also be that same alarm status. For example, if all NEs in a map have a NE alarm status of CRITICAL then the overall state of the map when considering the CRITICAL alarm status will also be CRITICAL as the services or functionality provided by the map will be critically affected. The same can be applied to any of the alarm statuses, e.g. if all NEs in a map have an alarm status of MINOR then the overall state of the map when considering the MINOR alarms will also be MINOR.

If the determined portion of the NEs with a particular NE alarm status is 0 then none of the NEs in the map have that particular NE alarm status and therefore the alarm state of the map for the particular NE alarm status being considered is NORMAL as the services or functionality provided by the map are not affected by the particular NE alarm status.

However, if the determined portion of the NEs with a particular NE alarm status lies between 1 and 0 then the alarm state of the map, for that particular NE alarm status, is determined based on the effect that the portion of NEs with a par- ticular NE alarm status will have on the services and functionality provided by the map taking into account the redundancy scheme and network architecture implemented by the network operator.

The determined portion of the NEs affected by a particular NE alarm status will lie between two of the predefined boundaries . The lower the portion of NEs with a particular NE alarm status then the impact on the services and functionality provided by the map is less as the remaining NEs in the map are still capable of providing those services and functionality. Therefore, depending on the portion of the NEs with a particular NE alarm status the alarm state of the map, in other words the alarm state of the services and functionality provided by the map, can be determined taking into ac- count the network architecture and redundancy scheme implemented by the network operator.

The predefined boundaries are ordered numerically starting from zero where boundary zero represents 1 and the final boundary, being boundary N represents 0. The alarm state of the map can be determined for a particular alarm status by reducing the severity of the NE alarm status being considered by the numerically ordered number of the upper boundary of the two boundaries between which the portion of NEs having the alarm status is positioned.

For example, if the alarm status being considered is CRITICAL and the network operator has set two boundaries then boundary zero (bθ) is 1, boundary one (bl) is the upper boundary set by the network operator, boundary two (b2) is the lower boundary set by the network operator and boundary three (b3) is 0. Also, as described hereinabove the alarm status for telecommunications as defined by ITU-T Rec. X.731/733, are an ordered list being CRITICAL, MAJOR, MINOR, WARNING and NORMAL. However, as a person skilled in the art will appreciate any number of boundaries may be set and any number of alarm statuses may be defined.

Accordingly, if a high portion of the NEs have the CRITICAL alarm status and the portion falls between bO and bl then the alarm state of the map for the NE CRITICAL alarm status can be determined by reducing the alarm status by the number of the upper boundary, which in this example is bO . Accordingly, the alarm state of the map for the NE CRITICAL alarm status in this example will be CRITICAL - 0 which is CRITICAL. Thus, if the portion of NEs having a CRITICAL alarm status is greater than bl then the services and func- tionality provided by the map will be critically affected.

If the portion of NEs having a CRITICAL alarm status is between the boundaries bl and b2 in this example then, as before, the alarm state of the map can be determined by reduc- ing the alarm status by the number of the upper boundary, which in this case is bl . Accordingly, the alarm state of the map for the NE CRITICAL alarm status will be CRITICAL - 1 which is MAJOR. Thus, if the portion of NEs having a CRITICAL alarm status is between bl and b2 then due to the network architecture and redundancy scheme the CRITICAL alarm status in some of the NEs can be compensated for without greatly affecting the services and functionality provided by the map. Therefore, the alarm state of the map can be re- duced from CRITICAL to MAJOR. The alarm state of the map is MAJOR as any further NEs developing a CRITICAL alarm status may critically affect the services and functionality provided by the map.

If the portion of NEs having a CRITICAL alarm status is between the boundaries b2 and b3 in this example then, as before, the alarm state of the map can be determined by reducing the alarm status by the number of the upper boundary, which in this case is b2. Accordingly, the alarm state of the map for the NE CRITICAL alarm status will be CRITICAL - 2 which is MINOR. Thus, if the portion of NEs having a CRITICAL alarm status is between b2 and b3 then due to the network architecture and redundancy scheme the CRITICAL alarm status in some of the NEs can be compensated for without af- fecting the services and functionality provided by the map. Therefore, the alarm state of the map can be reduced from CRITICAL to MINOR.

If the portion of NEs that have a CRITICAL alarm status is a value that falls directly on one of the predefined intermediate boundaries, e.g. a boundary that is not 1 or 0, then the predefined boundary on which the portion of the NEs with a CRITICAL alarm status falls is considered the lower boundary. For example, if the portion of NEs with a CRITICAL alarm status falls on the boundary b2 then this is considered the lower boundary and the alarm status will be reduced by 1 as the upper boundary in this case will be bl . However, as a skilled person in the art will appreciate, boundary on which the portion of the NEs with a CRITICAL alarm status falls could alternatively be considered the upper boundary.

In the above examples, depending on the boundaries between which the portion of NEs with a particular NE alarm status falls between the alarm state of the map for that particular NE alarm status can be determined as the NE alarm status being considered reduced by the numerically ordered number of the upper boundary. However, if any NE has the particular alarm status then the alarm state of the map for that particular NE alarm status cannot be reduced below a WARNING as the network operator should be informed that there is a fault in the map. Accordingly, the minimum the alarm state of the map for a particular NE alarm status where at least one NE has the NE alarm status will be WARNING.

For example, if the NE alarm status being considered is MINOR and if the portion of the NEs with a MINOR alarm status falls between the boundaries b2 and b3 in the above example, then the alarm state of the map would be MINOR - 2 which is

NORMAL. However, as a portion of NEs are affected by the MINOR alarm status then the map of NEs is not functioning normally and as such the alarm state of the map will be set at WARNING. Thus, the alarm state of a map for a particular NE alarm status will be the maximum of the alarm state determined and a WARNING alarm state.

As described hereinabove, if none of the NEs have a CRITICAL alarm status then the portion will be 0 and as such the alarm state of the map for the CRITICAL alarm status will be NORMAL .

In the above described example, the CRITICAL alarms were considered and depending on the portion of the NEs having a CRITICAL alarm status the alarm state of the map for the CRITICAL alarm status was determined.

Each of the further NE alarm statuses greater than NORMAL are also considered in turn individually and the alarm state of the map is determined for each NE alarm status. The overall alarm state of the map can then be determined as the highest severity alarm state of the alarm states determined for each of the NE alarm statuses.

The embodiments of the present invention can also be written in pseudo code. Firstly, N boundaries may be predefined that reflect the network architecture and redundancy scheme implemented in the telecommunication network. In particular, the boundaries are defined such that they reflect the impact on the services and functionality of a map M if a portion of NEs affected by a particular NE alarm status falls between two of the predefined boundaries. Each defined boundary bi, where i is an integer between 0 and N, lies between 1 and 0 where bO = 1 and bN = 0.

The network operator may set the boundaries to any value between 1 and 0 which reflect the impact that a portion of NEs having an NE alarm status will have on the services and func- tionality provided by the map taking into account the network architecture and the redundancy scheme implemented. The network operator may also set the boundaries such that a particular portion of NEs having a NE alarm status may be considered to have a greater or lower impact on the services and functionality provided by the map. For example, the network operator may set several boundaries either the same value or values that are very similar so that any portion lower or higher than the boundaries with the same value will reduce the NE alarm status being considered by different amounts. An overall alarm state of the map M may be determined as the highest severity SEV alarm state of the alarm states for the map determined for each NE alarm status. Each NE alarm status is considered in turn from severity SEV being CRITICAL to the severity SEV being WARNING and the portion P of NEs in the map M having each NE alarm status of severity SEV is used to determine an alarm state of the map for each of the NE alarm statuses.

Thus, for all NE alarm statuses SEV greater than or equal to WARNING an alarm state for the map M can be determined for each alarm status SEV as

Alarm State (M, SEV) = SEV, if P (M, SEV) = 1

Alarm State (M, SEV) = NORMAL, if P (M, SEV) = 0

Alarm State (M, SEV) = MAX (SEV - i, WARNING), if P (M, SEV) is in [bi, bi+1) and 0 < P (M, SEV) < 1

Once all of the alarm statuses have been considered then the overall alarm state of the map M can be determined as the max Alarm State (M, SEV) over all SEV with SEV > NORMAL.

Considering now a specific example as shown in Figure 4, the map 401 comprises fifteen NEs 402 which cooperate to provide particular services and functionality in a telecommunication network. The network operator may also implement any type of network architecture and redundancy scheme in order to provide resilience in the telecommunication network.

For the purpose of this example, we will firstly consider the CRITICAL NE alarm status and it is also assumed that the network architecture and redundancy scheme implemented in the map 401 will allow for several of the fifteen NEs 402 to be CRITICAL before the services and functionality are affected.

In this example, due to the network architecture and redun- dancy scheme implemented if twelve to fifteen NEs 402 have a CRITICAL failure then the services provided by the map 401 are seriously degraded and unacceptable. Therefore, the network will set a predetermined boundary bl as 0.8 which is the ratio of 12/15. Accordingly, if the portion of NEs with a CRITICAL alarm status is between b0 which is 1 (i.e. all fifteen NEs) and bl which is 0.8 (i.e. 12 NEs) then the alarm state of the map 401 is determined as CRITICAL - 0 (the CRITICAL NE alarm status being reduced by the number of the upper boundary which in this case is bθ) which is CRITICAL as corrective measures are urgently required as the services provided by the map 401 are unacceptable.

In this example, due to the network architecture and redundancy scheme implemented if six to eleven NEs 402 have a CRITICAL failure then the services provided by the map 401 may be degraded but may still be within the acceptable service limits as the remaining NEs 402 in the map can share the load and provide the services and functionality of the map 401. Therefore, the network operator will set the next pre- determined boundary b2 as 0.4 which is the ratio 6/15. Accordingly, if the portion of NEs 402 with a CRITICAL alarm status is less than but not including bl which is 0.8 (i.e. less than 12 NEs) and b2 which is 0.4 (i.e. 6 NEs) then the alarm state of the map is determined as CRITICAL - 1 (the CRITICAL NE alarm status being reduced by the number of the upper boundary which in this case is bl) which is MAJOR as corrective measures are required before the services and functionality provided by the map 401 are critically degraded. In this example, due to the network architecture and redundancy scheme implemented if three to five NEs 402 have a CRITICAL failure then again the services provided by the map 401 are not impacted or degraded as the remaining NEs 402 in the map 401 can share the load and provide the services and functionality of the map 401. Therefore, the network operator will set the next predetermined boundary b3 as 0.2 which is the ratio 3/15. Accordingly, if the portion of NEs 402 with a CRITICAL alarm status is less than but not including b2 which is 0.4 (i.e. less than 6 NEs) and b3 which is 0.2 (i.e. 3 NEs) then the alarm state of the map 401 for the CRITICAL NE alarm status is determined as CRITICAL - 2 (the CRITICAL alarm status being reduced by the number of the up- per boundary which in this case is b2) which is MINOR.

In this example, due to the network architecture and redundancy scheme implemented if one or two NEs 402 have a CRITICAL failure then the services provided by the map 401 are not affected and the remaining NEs 402 in the map 401 can share the load and provide the services and functionality of the map 401. Accordingly, if the portion of NEs 402 with a CRITICAL alarm status is less than but not including b3 which is 0.2 (i.e. less than 3 NEs) and greater than but not in- eluding b4 which is set at 0 then the alarm state of the map for the CRITICAL NE alarm status is determined as CRITICAL - 3 (the CRITICAL alarm status being reduced by the number of the upper boundary which in this case is b3) which is WARNING.

In the above example, if the portion of NEs is 1 (i.e. all of the NEs in the map) then the alarm state of the map 401 is equal to the alarm status being considered and as such the alarm state of the map is CRITICAL. If the portion of NEs is 0 (i.e. none of the NEs in the map) then the alarm state of the map will be NORMAL as none of the NEs in the map is affected by the CRITICAL alarm status.

Therefore, depending on the portion of the NEs 402 having a CRITICAL alarm status the alarm state of the map for the CRITICAL NE alarm status can be determined. This process is repeated for each of the NE alarm statuses MAJOR, MINOR and WARNING.

For example, if the portion of NEs having a MAJOR alarm status is 5 then the portion of NEs with a MAJOR alarm status is 5/15 which is 0.33 and falls between the predetermined boundaries b2 (which is in this example 0.4) and b3 (which in this example is 0.2) . Accordingly, the alarm state of the map for the MAJOR alarm status can be determined as MAJOR - 2 (which is MAJOR reduced by the number of the upper boundary b2) which is WARNING.

As described hereinabove, if any NEs have a particular NE alarm status then the alarm state of the map for that particular NE alarm status cannot be reduced lower than WARNING. For example, in the above example if only one of NEs in the map of fifteen NEs has a MINOR alarm status then the portion of NEs with a MINOR alarm status is 1/15 which is 0.066 and falls between the predetermined boundaries b3 (which in this example is 0.2) and b4 (which in this example is 0) . Accordingly, the alarm state of the map for the MINOR NE alarm status would be MINOR - 3 (which is MINOR reduced by the num- ber of the upper boundary b3) but this is not allowed as the network operator needs to be informed that the NE is not functioning correctly. Therefore, the alarm state of the map is determined as the maximum alarm state of the determined alarm state and WARNING. Accordingly, in this case the maxi- mum alarm state of MINOR - 3 and WARNING is WARNING and thus for the MINOR NE alarm status in this example the alarm state for the map would be determined to be WARING as this is the lowest allowable alarm state allowed when at least one NE has a particular NE alarm status .

Once each of the NE alarm statuses of CRITICAL, MAJOR, MINOR and WARNING have been considered and an alarm state for the map determined for each of the NE alarm statuses then the overall alarm state of the map can be determined as the maximum of the alarm states determined for each of the NE alarm statuses. The overall alarm state of the map may then be provided to the network operator as an indication of the state of the services and functionality provided by the par- ticular map.

As described above, the conventional systems have the disadvantage that if one NE in the fifteen NEs had a CRITICAL failure then the whole map would be considered CRITICAL even though the services and functionality provided by the map have not been degraded or impacted. Thus, the embodiments of the present invention have the advantage that the alarm state of the map is determined based on the alarm status of the NEs but in a manner that takes into consideration the network ar- chitecture and redundancy scheme implemented.

In the above example, the map contained fifteen NEs however a person skilled in the art will appreciate that the map can contain any number of NEs. Also the boundaries in the above example were chosen to reflect an arbitrary network architecture and redundancy scheme however a person skilled in the art will also appreciate that any number of boundaries can be predefined in order to reflect any network architecture and redundancy scheme that may be implemented by the network operator.

In the embodiments, the alarm state of a map may be calcu- lated periodically at time intervals required or set by the network operator. The alarm state of the map may also be calculated when the alarm state of the contained entities changes, e.g. when the alarm status of at least one of the NEs in the map changes. The alarm state may be calculated at either the periodic time intervals or when the alarm state of a contained entity changes or the alarm state may be calculated at both periodic intervals and when the alarm state of the contained entity changes .

The above described example is provided with reference to a particular redundancy scheme and network architecture where the network operator has set three intermediate boundaries being bl, b2 and b3. However, the alarm propagation mechanism described hereinabove can be used to determine the alarm status of a map which comprises any type of redundancy scheme and network architecture which is defined by any number of boundaries. This flexibility in the alarm propagation mechanism is achieved by the network operator setting any number of predefined boundaries at a value that accurately reflects the redundancy scheme and network architecture implemented by the network operator.

In other words, by defining the boundaries the network operator can decide the portion of entities contained in a par- ticular map which will need to have a particular alarm state for the map itself to have the same alarm state. For example, the network operator has the flexibility to define an upper boundary such that the map will have an alarm state of CRITICAL when a particular portion of the contained entities is CRITICAL.

Accordingly, the embodiments advantageously enable a flexible alarm propagation mechanism which adjusts the alarm state of a map of contained entities based on the portion of the contained entities with a particular alarm state.

While preferred embodiments of the invention have been shown and described, it will be understood that such embodiments are described by way of example only. Numerous variations, changes and substitutions will occur to those skilled in the art without departing from the scope of the present invention as defined by the appended claims. Accordingly, it is in- tended that the following claims cover all such variations or equivalents that fall within the spirit and scope if the invention .

Claims

1 . A method compri sing : determining for each of a set of network element alarm statuses a portion of network elements in a map of network elements having said network element alarm status; determining an alarm state of said map for each of said network element alarm statuses based on said portion of network elements having said network element alarm status; and determining an overall alarm state of said map as the maximum severity alarm state of the alarm states determined for said map for each network element alarm status.

2. The method as claimed in claim 1, in which said step of determining said alarm state of said map for each of said network element alarm statuses further comprises: determining if said portion of network elements in said map with said network element alarm status is one and setting said alarm state for said map for said network element alarm status equal to said network element alarm status.

3. The method as claimed in claim 1 or 2, in which said step of determining said alarm state of said map for each of said network element alarm statuses further comprises: determining if said portion of network elements in said map with said network element alarm status is zero and setting said alarm state for said map for said network element alarm status equal to NORMAL.

4. The method as claimed in any one of the preceding claims in which said step of determining said alarm state of said map for each of said network element alarm statuses further comprises : identifying two predetermined boundaries between which said portion of network elements in said map with said network element alarm status is located; and setting said alarm state for said map for said network element alarm status equal to the network element alarm status being reduced by the numerically ordered number of an upper boundary of the two boundaries between which said portion of network elements in said map with said network element alarm status is located.

5. The method as claimed in claim 4 in which said method further comprises: setting a set of boundaries between 1 and 0 wherein said boundaries define a network architecture and redundancy scheme.

6. The method as claimed in claim 5 in which said set of boundaries are numerically ordered starting with bO representing 1 and ending with boundary bN representing 0.

7. The method as claimed in any one of the preceding claims in which said portion is determined as the ratio of the network elements having said network element alarm status to the total number of network elements in said map.

8. A server configured to: determine for each of a set of network element alarm statuses a portion of network elements in a map of network elements having said network element alarm status; determine an alarm state of said map for each of said network element alarm statuses based on said portion of network elements having said network element alarm status; and determine an overall alarm state of said map as the maximum severity alarm state determined for said map for each network element alarm status .

9. A computer program product comprising computer readable executable code for: determining for each of a set of network element alarm statuses a portion of network elements in a map of network elements having said network element alarm status; determining an alarm state of said map for each of said network element alarm statuses based on said portion of network elements having said network element alarm status; and determining an overall alarm state of said map as the maximum severity alarm state determined for said map for each network element alarm status.