WO2007045637A1

WO2007045637A1 - Method for reporting alarms of a network element

Info

Publication number: WO2007045637A1
Application number: PCT/EP2006/067460
Authority: WO
Inventors: Sergio Manuel Silva
Original assignee: Nokia Siemens Networks Gmbh & Co. Kg
Priority date: 2005-10-21
Filing date: 2006-10-16
Publication date: 2007-04-26

Abstract

Only an alarm signal (AS1 - ASn), which persists for a predetermined time, is leading to a primary alarm message (AM1 - AMn), similar primary alarm messages of the network element are combined to reduce the number of alarm messages. The alarm messages are analysed and a network alarm message (NE-AM) indicating a primary alarm cause is forwarded to a management system.

Description

Method for reporting alarms of a network element

The invention refers to a method for reporting alarms of a network element.

BACKGROUND OF THE INVENTION

Primary alarms in a network element can lead to a set of secondary alarms . A primary alarm is derived from a primary fault condition (root cause), e.g. the interruption of a transmitting line or a fault in a line card. A primary fault alarm can produce a flood of secondary alarms that overload an upper level management system.

Up to now, the handling of alarm messages was done by a high level management system, which was responsible e.g. for alarm analysis and for suppressing of secondary alarms.

Such a system is described in the patent US 5,388,189. All alarms received by a network manager are transferred to an event manager, queued and than filtered by an alarm filter to remove redundant alarms. The result is an immense number of alarm messages between the network elements and the upper level management system.

SUMMARY OF THE INVENTION

An object of the invention is to provide an improved alarm signalling method.

This object is met by the method defined by the method steps in an independent claim 1. Option method steps are defined in dependent claims . Besides filtering of the alarm signals the reduction of a plurality of alarm messages is a main advantage of the invention. For a plurality of similar alarm signals only one network element alarm message is forwarded from a network element to the management system. The communication between the network element and the management system is reduced significantly and therefore the overload possibility is avoided.

Another important advantage is the preference of primary alarm messages. These messages are derived from primary (basic) fault conditions, which make the fault analysis much easier. Secondary alarm signals, which can by identified as a result of a primary alarm, can be suppressed by the network element or stored and transmitted later according to the wishes of an operator. Therefore a misinterpretation of an alarm and root cause is significantly reduced. The better information about the root cause that lies behind the (primary) failure message results in operation cost reduction.

Fault detection is facilitated when instead of reporting an alarm message and supplement information the measured or assumed fault condition can be reported to the management system/operator. A primary fault condition is identified automatically and forwarded in addition to a primary alarm signal or as a part of the network element alarm message.

New inserted cards in a network element are adapted automatically without the help of an operator.

DESCRIPTION

An embodiment of the present invention will now be described, by way of example only, with reference to the attached drawings wherein Fig.l shows a block diagram schematically illustrating a network element and the principle of alarm generation,

Fig 2 shows a block diagram of a preferred embodiment of the invention, and

Fig.3 shows a time diagram of corresponding alarm messages.

DESCRIPTION OF PREFERRED EMBODIMENTS

Figure 1 shows a simplified embodiment of a network element with a plurality of line cards LCl - LCn and at least a processor cards PCl - PCn (at least one) . Each line card has an line input Il - In and an line output Ol - On for receiving and transmitting information, and also connecting points CIl - CIn, COl- COn for exchanging data e.g. over a data bus BU between the line cards LCl - LCn and the processor cards PCl - PCn.

The line inputs and the line outputs and the connecting points in this embodiment are fed to a fist Alarm Detection Unit ADUl (not all possibilities are shone) . In addition other monitoring signals MLl - MLn of the line cards are fed to the ADUl. The monitoring signals MPl - MPn of the processor cards are supervised e.g. by a second alarm detection unit ADU2. The alarm detection units generate 'primary alarm messages' (AMI, AMn; Fig. 3) . A first total primary alarm message AMSl is derived from the primary alarm messages, which are derived from the line input and the monitoring signals MLl - MLn, and a second total primary alarm message AMS2 is derived from the monitoring signals MLl - MPn. Both are fed to an 'alarm analysing unit' AAU, which generates a 'network element alarm message' NE-AM, containing the primary alarm messages derived from the monitoring signals by the alarm detection units ADUl, ADU2 in a compressed data format. Now let us assume e.g. a cable for receiving and transmitting all line signals is broken and no signals are received from this time on. The input signals are fed to analogue or digital filters FI of the alarm detection units. A special filter is provided for each alarm cause. The filtering mechanism may use a "direct acyclic graph" or a multi path direct acyclic graph.

According to Fig.3 the filters FI receive fault conditions "no signal" from the line terminals II, In at an instant t_Fi, respectively t_Fn (dotted lines changing from logical 0 to logical 1), identifies the fault conditions as "no signal" and generates at a time instant t^n, respectively tAMn alarm signals ASl, ASn (logical 1 of the solid lines) . Each alarm signal has to remain for a certain time before a ^Λprimary alarm message' AMI, AMn is generated at tAMi respectively tAMn • In this embodiment the filtering is done by the alarm detection units ADUl, ADU2 themselves. It could also be done by the alarm analysing unit AAU. All these alarm messages as a whole derived from the line cards and the processor cards are referred to as alarm messages AMSl and AMS2.

In a special embodiment all signalling is done by a clock signal, whose effective slopes are labelled by time marks t_o - t3. Each alarm signal has to be stable for at least one integration period ΔT, that means between two effective slopes of the clock signal between e.g.ti and t₃. Instead at the time instants tAMi and tAMn the primary alarm messages AMI and AMn are generated (respectively output for further processing by the alarm analysing unit) at t2 and combined to total primary alarm messages AMSl, AMS2 and to only one "Network Element Alarm Message" NE-AM (or at t₃) . This combination can be done by the alarm detection units ADUl - ADUn or the alarm analysing unit AAU. A primary fault condition is identified automatically. When an alarm condition is detected, the line card, the ADU or the AAU translate this condition to a catalogued cause. A corresponding digital alarm message as a part of the network element alarm message NE-AM is forwarded from the AAU to a management system indicating "Line Interruption Il -In" . An alarm message and also different alarm messages can be forwarded in addition to the filtered primary alarm signals.

The interruption of the transmission lines results in further "secondary" alarm messages. At the connecting points CIl - CIn no signals are outputted and e.g. the processor cards PCl - PC3 do not receive valid data signals. So the fault leads to secondary failure conditions and to "secondary alarm messages" generated by the alarm detection units ADUl and ADU2. The alarm analysing unit AAU monitors all these alarm messages and according to a primary (root/basic) cause detection algorithm the secondary alarms are recognised and suppressed or stored and forwarded on request of an operator or when the connection (traffic condition) to the management system allows it.

Because in the assumed failure case all lines are interrupted, the voltages at the line outputs Ol - On are too high and a second network element alarm message is generated indicating "Line Interruption 01-On" . Another possibility is to generate a "combined" network element alarm message "Line Interruption H-In; 01- On".

According to an preferred embodiment of the invention shown in Fig. 2 the alarm detection, the generation of alarm messages, and in some cases the "primary alarm detection" (route fault detection) , and also the translation from a detected or identified (primary or secondary) alarm to the associated fault (root cause) is realized on each line card LCl-LCn and processor card PCl-PCn. All alarm messages AMSl from the line cards and AMS2 from the processor cards are transmitted to the alarm analysing unit AAU.

In this case only a "supervisory" alarm analysis and the generation of the network element alarm signal may be done by the AAU. This arrangement makes changes in the network element, e.g. the insertion of new or additional line cards much easier because a new card generates its own alarm messages .

Claims

1. Method for reporting alarms of a network element, characterised in that an alarm signal (ASl - ASn) is filtered and only an alarm signal (ASl - ASn) , which persists for a predetermined time, is leading to a primary alarm message (AMI, AMSn) and similar primary alarm messages (AMI, AMn) of the network element are combined to a network element alarm message (NE- AM) and forwarded to a management system.

2. Method for reporting alarms according to claim 1, characterised in that only primary alarm messages (AMI, AMn) as a result of primary fault condition are forwarded as network element alarm messages (NE-AM) .

3. Method for reporting alarms according to claim 1, characterised in that secondary alarm messages as a result of primary fault conditions are suppressed or stored and forwarded later as secondary network element alarm messages with lower priority than primary alarm messages (AMI, AMn) .

4. Method for reporting alarms according to claim 1-3, characterised in that a primary fault condition is identified automatically and forwarded in addition to a primary alarm message (AMI, AMn), a total primary alarm message (MSl, MS2) or as a part of the network element alarm message (NE-AM) .

5. Method for reporting alarms according to claim 1, characterised in that primary alarm messages (AMI, AMn), total primary alarm messages (AMSl, AMS2), and network element alarm messages (NE-AM) are derived and forwarded with a system clock signal.

6. Method for reporting alarms according to claim 1, characterised in that a card (LCl - LCn, ...) inserted into that network element generates its own set of alarm signals or alarm messages automatically according to its functions .

7. Method for reporting alarms according to claim 1 -6, characterised in that the primary alarm message (AMI, AMn) is analysed and the network alarm message (NE-AM) indicating a primary alarm cause is forwarded to the management system.

8. Method for reporting alarms according to claim 1, characterised in that a card (LCl - LCn, ...) determines for each detected secondary alarm which primary alarm or which primary alarm cause was reasonable.

9. Method for reporting alarms according to claim 8, characterised in that a line card (LCl - LCn) or a processor card (PCl - PCn) forwards the primary alarm message or the primary alarm cause to am alarm analysing unit (AAU) .