J. F. Jordaan and M. E. Paterok. Event correlation in heterogeneous networks using the OSI management framework. In H. G. Hegering and Y. Yemini, editors, Integrated Network Management, III, pages 683--695. Elsevier Science Publishers B.V., 1993.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the current network state (evaluates to true) the action in the rule s right hand side is taken. Actions can either assert or clear correlations. Correlation assertions are meta events that can in turn be matched in the left hand sides of other rules. The event correlation system describe in [16] automates certain aspects of the problem by exploring hypotheses about the relationships between events to form correlations and to infer causality. In particular, events 9 are hypothesized to be related if the objects generating the events are adjacent in the protocol stack, if they are at ....

....sense that they are predictiveoformay be predicted by other portions of network state. Our approach differs from most current approaches to event correlation in that we consider state information obtained via time driven monitoring rather than just event driven monitoring. The work described in [16] considers more detailed state information, but only after particular events are hypothesized to be correlated. The former information allows reasoning about the nature of existing problems as manifested byvarious events# the latter information allows reasoning about the nature of problems that ....

J. F. Jordaan and M. E. Paterok. Event correlation in heterogeneous networks using the OSI management framework. In H. G. Hegering and Y. Yemini, editors, Integrated Network Management, III, pages 683--695. Elsevier Science Publishers B.V., 1993.

....systems are well suited for learning correlation patterns. They are resilient to changes in network configuration. However, case based systems require an application specific model for the resolution process and are computationally complex. Model traversing techniques Model traversing techniques [11, 15, 24, 26] use a formal representation of a communication system with clearly marked relationships between network entities. Failures in the communication system propagate along relationships between network entities. By exploring these relationships, starting from the network entity that reported an alarm, ....

J. F. Jordaan and M. E. Paterok. Event correlation in heterogeneous networks using the OSI management framework. In IM'93 [19], pp. 683--695.

.... values Threshold violations and system bottleneck prediction Support for uncertainty and lost events 13 6 Related work In the past various event correlation techniques were proposed including rule based systems [18, 26] model based reasoning systems [13, 21] model traversing techniques [14, 15], case based systems [17] fault propagation models [9, 16] and the code book approach [27] Rule based systems are composed of rules (productions) of the form conclusion if condition. The condition part is a logical combination of propositions about the current set of received alarms and the ....

....unwanted rule side effects and decrease the overall number of correlation rules. Another group of approaches incorporate an explicit representation of the structure and function of the system being diagnosed. The representation provides information about dependencies between network components [11, 13, 14, 15, 16] or about cause effect relationships between network events [9, 21] The fault isolation process explores the network model to verify correlation between events. Model based reasoning systems [13, 21] utilize inference engines controlled by a set of correlation rules, which contain model ....

[Article contains additional citation context not shown here]

J. F. Jordaan and M. E. Paterok. Event correlation in heterogeneous networks using the OSI management framework. In H. G. Hegering and Y. Yemini, editors, Integrated Network Management III, pages 683--695, San Francisco, CA, Apr. 1993. North-Holland.

.... Graph size one fault two faults three faults four faults Figure 9: Fault localization time with Algorithm 7 Related work In the past, various event correlation techniques were proposed including rule based systems [22, 34] model based reasoning systems [15, 24] model traversing techniques [16], case based systems [21] fault propagation models [12, 18] and the codebook approach [35] Most of the above approaches utilize deterministic reasoning. This paper focuses on non deterministic event correlation which is unavoidable in fault diagnosis related to quality of service degradation ....

J. F. Jordaan and M. E. Paterok. Event correlation in heterogeneous networks using the OSI management framework. In Hegering and Yemini [14], pp. 683--695.

....nondeterministic fault localization, yielding high accuracy even for approximate probability input data. IX. RELATED WORK In the past, various event correlation techniques were proposed including rule based systems [22] 23] model based reasoning systems [24] 25] model traversing techniques [26], case based systems [27] fault propagation models [6] 7] and 0 7803 7476 2 02 17.00 (c) 2002 IEEE. 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1 rate Network size exact three confidence levels two confidence levels one confidence level Fig. 10. Detection rate of Algorithm 1 ....

J. F. Jordaan and M. E. Paterok, "Event correlation in heterogeneous networks using the OSI management framework," In Hegering and Yemini

....network may result in a large number of alarms, and it is often very difficult to isolate the true cause of the fault. The problem becomes even worse when several faults occur coincidentally in a telecommunications network. Recently, a number of researchers have proposed algorithms [1, 2, 5, 6, 7, 3, 4] to perform alarm correlation and fault identification. Some of these techniques are even in the advanced prototyping phase [3] However, all the proposed methods presume that the telecommunications network is managed by a single management center, that has a global view of the network and access ....

....becomes even worse when several faults occur coincidentally in a telecommunications network. Recently, a number of researchers have proposed algorithms [1, 2, 5, 6, 7, 3, 4] to perform alarm correlation and fault identification. Some of these techniques are even in the advanced prototyping phase [3]. However, all the proposed methods presume that the telecommunications network is managed by a single management center, that has a global view of the network and access to all management data. Telecommunications networks are increasingly partitioned into distinct management domains, each managed ....

J.F.Jordaan and M.E.Paterok. Event Correlation in Heterogeneous Networks Using the OSI Management Framework. Third IEEE/IFIP International Symposium on Integrated Network Management, San Francisco,, April 18-23 1993.

....Most of the proposed algorithms provide solutions for physical layer fault localization.In [1] a general estimation technique was proposed using information cost which is applicable to any type of fault. An application of [1] with emphasis on the physical layer failures, was presented in [2] In [6] an application using the OSI fault management framework and ideas from [2] was described. Our objective is to define a general fault localization framework and define algorithms applicable to both physical layer failures and failures at any layer in the protocol stack. The problem can be ....

J.F.Jordaan and M.E.Paterok. Event Correlation in Heterogeneous Networks Using the OSI Management Framework. Third IEEE/IFIP International Symposium on Integrated Network Management, San Francisco, , April 18-23 1993.

....These questions suggest that the design of suitable management services should consider the representation and reasoning techniques used by knowledge based management systems. Among such systems, model based systems have gained a rapidly increasing acceptance for use in event and alarm correlation [4], 5] 6] as well as for fault and configuration management of distributed services and applications [7] 8] The latter category poses new challenges, as there are no standard models at the service layer. Therefore, management tools should provide flexible and powerful modeling schemes along ....

....map to, mirror, and respond to the managed system. Once these models have been constructed, they are used as the basis of on line tools that monitor operating networks to detect and diagnose faults as they occur. The solution to event correlation and fault isolation problems presented in [4] and [5] is based on modeling network resources as objects in a generic object model, and on developing relationships between them. The relationships model links across layers of the protocol stack, and between connections at the same layer. To correlate the network objects reporting fault events ....

[Article contains additional citation context not shown here]

J. Jordaan and M. Paterok, "Event correlation in heterogeneous networks using the osi management framework," in Proceedings of the Third IFIP/IEEE International Symposium on Integrated Network Management, pp. 683--695, 1993.

....the current network state (evaluates to true) the action in the rule s right hand side is taken. Actions can either assert or clear correlations. Correlation assertions are meta events that can in turn be matched in the left hand sides of other rules. The event correlation system describe in [16] automates certain aspects of the problem by exploring hypotheses about the relationships between events to form correlations and to infer causality. In particular, events are hypothesized to be related if the objects generating the events are adjacent in the protocol stack, if they are at ....

....that they are predictive of or may be predicted by other portions of network state. Our approach differs from most current approaches to event correlation in that we consider state information obtained via time driven monitoring rather than just event driven monitoring. The work described in [16] considers more detailed state information, but only after particular events are hypothesized to be correlated. The former information allows reasoning about the nature of existing problems as manifested by various events; the latter information allows reasoning about the nature of problems that ....

J. F. Jordaan and M. E. Paterok. Event correlation in heterogeneous networks using the OSI management framework. In H. G. Hegering and Y. Yemini, editors, Integrated Network Management, III, pages 683--695. Elsevier Science Publishers B.V., 1993.

....of the system on the experience gained with the system itself. The expert knowledge, which forms the empirical associations collected into rules in a RBR system, or the similarity metrics and adaptation functions in a CBR system, is now formalized into the model of the system to be managed. (Jordaan and Paterok 1993) describes a prototype event correlation application which needs very little or no preconfigured knowledge , compared to RBR systems. The underlying idea is that in practice almost all objects modeling the network are related in some fashion, but just a few relationships prove to describe fault ....

....the network are related in some fashion, but just a few relationships prove to describe fault propagation effectively, and cover the vast majority of, otherwise ad hoc, heuristic associations. Keeping modeling simple is the underlying idea in (Crawford et al. 1995) where the approach outlined in (Jordaan and Paterok 1993) is further formalized. A distinct modeling technique that has recently emerged in the field of network management is the utilization of constraints. However, the constraint based technique has not been explicitly related to the existing work on constraint satisfaction problems (CSPs) as they are ....

Jordaan, J.F. and Paterok, M.E. (1993) Event correlation in heterogeneous networks using the OSI management framework. In H.-G. Hegering and Y. Yemini, editors, Integrated Network Management, III, 683-695. Elsevier Science Publishers B.V., North-Holland.

....Most of the proposed algorithms provide solutions for physical layer fault localization.In [2] a general estimation technique was proposed using information cost which is applicable to any type of fault. An application of [2] with emphasis on the physical layer failures, was presented in [1] In [7] an application using the OSI fault management framework and ideas from [1] was described. Our objective is to define a general fault localization framework and define algorithms applicable to both physical layer failures and failures at any layer in the protocol stack. The problem can be ....

J.F.Jordaan and M.E.Paterok, " Event Correlation in Heterogeneous Networks Using the OSI Management Framework," Third IEEE/IFIP International Symposium on Integrated Network Management, San Francisco,, April 18-23 1993.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC