I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE/ACMTransactions on Networking, 3:753--764, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....Since fault propagation generates alarm propagation, techniques for alarm correlation are relevant to the third step in our program detecting activation of a generic propagation mechanism. There is a large body of literature on alarm correlation. Some examples include [JW93] JW95] [KS95], N95] YK 96] HSV99] among others. We also note that in the MAGDA Project [FBJ 00] the interaction of network elements in fault propagation is modeled using Petri nets. In this paper, we focus on a single example of a generic propagation mechanism: overload propagation in the control plane ....

I. Katzela and M. Schwartz, "Schemes for fault identification in communication networks," IEEE Transactions on Networking, 3 (6), 1995.

....and positive symptoms in the analysis. As shown in a simulation study, the technique is resilient both to noise in the symptom data and to the inaccuracies of the probabilistic fault propagation model. 1. Introduction This paper presents a non deterministic event driven fault localization [9, 10, 17] technique, which uses a probabilistic symptom fault map as a fault propagation model. While investigating fault localization techniques suitable for bipartite fault propagation models, this paper states the following objectives: Usage of probabilistic reasoning, which is necessary to diagnose ....

.... suitable for bipartite fault propagation models, this paper states the following objectives: Usage of probabilistic reasoning, which is necessary to diagnose Byzantine problems or when relationships among system events may not be determined with certainty, e.g. due to their dynamic nature [5, 6, 8, 10, 11]. Ability to isolate multiple simultaneous faults even if their symptoms overlap [6, 10] which improves the technique s applicability to large systems. Event driven diagnosis, which avoids the inflexibility of window based tools [1] is not prone to inaccuracies resulting from an ....

[Article contains additional citation context not shown here]

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE Transactions on Networking, 3(6):733--764, 1995.

....a single fault in the network will cause a flood of alarm signals emanating from different interrelated network elements. Knowledge of element interconnections is essential to filter out secondary alarm signals and correlate primary alarms to pinpoint the original source of failure in the network [1], 2] Furthermore, a full physical map of the network enables a proactive analysis of the impact of link and device failures. Despite the critical role of physical topology information in enhancing the manageability of modern IP networks, obtaining such information is a very difficult task. The ....

I. Katzela and M. Schwarz, "Schemes for Fault Identification in Communication Networks", IEEE/ACM Transactions on Networking,vol.3, no. 6, pp. 753--764, Dec. 1995.

....on the typical, packet level and end to end behavior (a notable exception being [7] Our focus is mainly at the flow level and on identifying frequency characteristics of anomalous network traffic. There have been many prior studies of network fault detection methods. Example include [8] [9], 10] Feather et al. use statistical deviations from normal traffic behavior to identify faults [11] while a method of identifying faults by applying thresholds in time series models of network traffic is developed in [12] These studies focus on accurate detection of deviations from normal ....

I. Katzela and M. Schwartz, "Schemes for fault identification in communications networks," IEEE/ACM Transactions on Networking, vol. 3(6), pp. 753--764, December 1995.

....to be very affective. The work in [20] can localize a fault to a portion of the multicast tree that is between two branch points. Similarly, our approach can localize performance problems between two branching points on the network topology. Dependency graphs are commonly used in fault isolation [15], 24] These approaches are based on first forming a dependency graph that models the propagation of faults in the network and then searching this graph to isolate the smallest set of nodes that are the likely culprits. There are parallels between the approach we used in this paper to localize ....

....and hence is the set to which we can suspect for poor call quality. For the example, this computation results with which is a reduction from the initial suspicious set of size 5. 13 The approach above is similar to the fault dependency model graph based techniques such as those in [15] and [24] if the dependency graph is formed as follows: The dependency graph is directed and bipartite where the nodes denote both individual links and end to end paths. There is a directed edge in the dependency graph from each node denoting an end toend path to each node denoting a link on the ....

I. Katzela and M. Schwartz. "Schemes for fault identification in communication networks,", IEEE/ACM Transactions on Networking, 3(6):753--764, 1995. 14

....knowledge among multiple, hierarchically organized managers. The cooperation among managers is illustrated with examples, and the results of a preliminary performance study are presented. 1 Introduction End to end network service failure diagnosis [15, 17] is a sub task of fault localization [7, 9, 19] that isolates host to host services responsible for availability or performance problems associated with a communication between end hosts. In [16, 17] probabilistic inference was applied to provide a non deterministic solution to this problem. The probabilistic fault propagation model ....

....fault model, in which parentless vertices (root causes) are labeled with independent failure probabilities and the directed edges are weighted with conditional probabilities representing the strengths of causal influences between host to host and end to end service failures. Typically [9, 10, 17], a noisy OR probability model [12] is used. Diagnosing end to end service failures may be mapped into the problem of finding the most probable explanation of the observed evidence in belief networks [12] While the problem is known to be NP hard [2] a polynomial time inference algorithm was ....

[Article contains additional citation context not shown here]

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE Transactions on Networking, 3(6):733--764, 1995.

....argued that typical networked systems provide plenty of information necessary to infer the existence of faults [41] The following paragraphs present some common problems that have to be addressed by a fault localization technique. Fault evidence may be ambiguous, inconsistent and incomplete [8, 14, 28]. Ambiguity in the observed alarm set stems from the fact that the same alarm may be generated as an indication of many different faults. Inconsistency is a result of a disagreement between different devices with regard to the facts related to network operation; one object may have the perception ....

....and unrelated faults may happen within a short time of one another. The event management system should be able to detect unrelated simultaneous problems even if they generate overlapping sets of alarms. Multiple different fault hypotheses may exist that explain the set of observed symptoms [28]. Given that a single alarm may indicate different types of faults that occurred in different communication devices, fault localization may be unable to give a definite answer. Some approaches discussed in Section 3 combine fault localization with testing in order to resolve these ambiguities. ....

[Article contains additional citation context not shown here]

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE Transactions on Networking, 3(6), 1995.

.... 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 system state [18, 26] the conclusion determines the state of ....

....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]

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE Transactions on Networking, 3(6), 1995.

....network, and then the approximations based on Pearl s algorithms and exact bucket tree elimination algorithm are designed and evaluated through extensive simulation study. Keywords Fault localization, probabilistic inference, belief networks, event correlation 1 Introduction Fault localization [15, 18, 35], a central aspect of network fault management, isolates the most probable set of root problems based on their external manifestations, called Prepared through collaborative participation in the Communications and Networks Consortium sponsored by the U. S. Army Research Laboratory under the ....

....for endto end service failures in a given layer. The proposed solutions allow the management system to perform fault localization iteratively in real time. In the past, fault diagnosis efforts concentrated mostly on detecting, isolating, and correcting faults related to network connectivity [9, 18, 33, 35]. The diagnosis focused on lower layers of the protocol stack (physical and data link layers) 24, 35] and its major goal was to isolate faults related to the availability of network resources, such as broken cable, inactive interface, etc. Modern enterprise environments increasingly demand ....

[Article contains additional citation context not shown here]

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE Transactions on Networking, 3(6):733--764, 1995.

....more likely than x 1 failures [MTB99, MT00, KW87, YS93] etc. 4. the quantity of memory they use: large memory requirements [HE93] to store failure history, or more modest memory requirements [LR95, BHS93, GH97] to store for example an FSM or neural network parameters. 5. intolerance [HE93, KS95] or tolerance [GH97, MTB99, MT00] to false and missing alarms. 3 Approaches to the fault location problem All techniques performing fault diagnosis rely on the analysis of symptoms and events that are generated during the occurrence of the fault. We divide them in two main categories. The first ....

....incomplete information, ffl they do not require learning, ffl they can cope with unforeseen failures The drawback is the difficulty of developing good model for complex networks. Some works belonging to this approach are: 1(a) probabilistic reasoning systems developed by Katzela and Schwartz [KS95] and Wang and Schwartz [WS93] The input parameters of these systems are failure probabilities, which may be difficult to be determine for optical networks because of their time dependence. These systems also take as input the conditional probabilities of failure of an element given the failure ....

[Article contains additional citation context not shown here]

I. Katzela and M. Schwartz. Schemes for Fault Identification in Communication Networks. IEEE/ACM Transactions on Networking, 3(6), December 1995.

....utilized for the networks operating in an unreliable environment such as wireless and or military networks. 1 INTRODUCTION To improve the network ability to provide reliable services to end systems, a management system needs to efficiently and accurately identify the occurring network failures [13, 25]. A common procedure is to correlate network or service layer symptoms; however, this process is usually impaired by the large number of a system s layers and parameters [1, 8, 22] their interactions, and the uncertainty about their state. This paper presents a preliminary study of applying ....

....confidence level their impact on the accuracy of symptom correlation. The elements that account for the highest variations of the correlation accuracy are selected as probable failure points. Most of the existing techniques assume that the existence of multiple simultaneous faults is negligible [13]. Such an assumption is justified only for networks operating in a reliable environment. On the other hand, soft link failures due to jamming, misbehaving nodes, or difficult weather conditions are not likely to be limited to a single link. Similarly, battlefield or other military applications are ....

[Article contains additional citation context not shown here]

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE Trans. Networking, 3(6), 1995.

....be used to describe a wide range of fault localization problems. The paper expands on our previous research on applying belief networks techniques to fault localization [5] Section IV) by taking into account comprehensive information about the system behavior. Most previous work on this subject [6], 7] performs fault localization based solely on the information about malfunctioning system components (i.e. negative symptoms) In this paper, we show that positive information, i.e. the lack of any disorder in some system components, may be used to improve the accuracy of fault localization ....

....Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation thereon. state is frequently disturbed by the presence of lost and or spurious symptoms (usually referred to as observation noise) Although many researchers have suggested [6], 8] that fault localization should be resilient to the existence of spurious and or lost alarms, we are aware of only one technique [8] that incorporates lost and spurious symptoms into deterministic fault localization. In this paper, we propose a technique that allows lost and spurious symptoms ....

[Article contains additional citation context not shown here]

I. Katzela and M. Schwartz, "Schemes for fault identification in communication networks," IEEE Transactions on Networking, vol. 3, no. 6, pp. 733--764, 1995.

....of analysis can be used to discover the source(s) of attacks thus complying with the first goal. Moreover, it can analyze a wide range of timevarying attacks in accordance with the robustness goal. A good treatment of alarm correlation for diagnostic of failures and of related work can be found in [21]. A general treatment of the Bayesian analysis involved in diagnostic problem solving is given in [34, 35] All of the above results assume some degree of centralized processing of alarms, usually at the network and subnetwork level. The time complexity of the software processing grows in some ....

I. Katzela amd M. Schwartz, "Schemes for Fault Identification in Communication Networks," IEEE/ACM Transactions on Networking, vol. 3, no. 6, December 1995, pp. 753-764.

....in legacy systems. They can also be adapted to different types of devices. As security needs change, wrappers can be added, upgraded or replaced. We do not address the issue of how to correlate the alarms due to attacks. The problem of identifying failures using alarms has been examined in [6,39]. There are several issues concerning the interpretation of and the reaction to alarms in very high speed networks. In this section and the following section, we present and analyze the performance of a novel approach for QoS monitoring to guard against denial of service attacks. We often refer ....

I. Katzela and M. Schwartz, Schemes for fault identification in communication networks, IEEE/ACM Transactions on Networking 3(6) (1995), 753--764. M. Mdard et al. / Node wrappers for QoS monitoring in transparent optical nodes 267

....true for passive elements such as optical fibers. The algorithm supports: i) multiple failures and (ii) passive network elements that do not generate alarms but may fail. In the literature there are algorithms for the localization of failures. Some of them work with probabilities of failure [1]. The problem of working with probabilities is that they are changing continuously depending on external parameters (for example age, temperature and pressure) Other algorithms use Management Information Bases to look for information related with the alarms [2] The proposed algorithm tries to ....

I. Katzela and M. Schwartz, "Schemes for Fault Identification in Communication Networks," IEEE/ACM Transactions on Networking, vol. 3, December 1995.

.... Intelligence suffer from being dependent on prior knowledge about the fault conditions on the network and the rules developed do not adapt well to a changing network environment [9] Research in the field of fault management is primarily focussed on fault localization by alarm correlation [20][14]. Fault localization is usually based on the assumption that the alarms provided by the network nodes are true and the relevance of temporal information in the alarms is ignored or assumed accurate. The issue of temporal information has been partially addressed [2] by introducing time correlated ....

I. Katzela and M. Schwarz. Schemes for fault identification in communication networks. IEEE/ACM Trans.Networking, 3:753--764, 1995.

....a single fault in the network will cause a flood of alarm signals emanating from different interrelated network elements. Knowledge of element interconnections is essential to filter out secondary alarm signals and correlate primary alarms to pinpoint the original source of failure in the network [1], 2] Furthermore, a full physical map of the network enables a proactive analysis of the impact of link and device failures. Early identification of single points of failure that could disrupt a large fraction of the user community allows the network manager to improve the survivability of the ....

Irene Katzela and Mischa Schwarz, ""Schemes for Fault Identification in Communication Networks"," IEEE/ACM Transactions on Networking, vol. 3, no. 6, pp. 753--764, Dec. 1995.

....not specific to event correlation. Many management applications need it or benefit from it, e.g. to compute service availability like in [DRK97] for change management or network planning. Among the extensions of dependency graphs are those that associate probabilities to the arcs of the graph ( KS95] Thus the demand is not to invent a new model for event correlation but to formalize one that is needed anyway so that it is available to tools like an event correlator. 5 3.2 Doing Event Correlation with Dependency Graphs Given a dependency graph, our event correlator works as depicted in ....

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. 3(6):753--764, December 1995.

....that take a detailed view to all network entities. By contrast, in fault localization probabilistic models are used that look at the system from a global perspective. A number of algorithms that use probabilistic reasoning for localizing failed network components appeared in the literature [4] [9]. In this paper, we developed a framework for a class of fault localization and identification problems that can be modeled using belief networks [10, 11] When a fault occurs and the problem is reported to the fault manager (either through the triggering of an alarm or a by a report from the ....

I. Katzela and M. Schwartz, "Schemes for fault identification in communication networks, " IEEE/ACM Trans. Networking, Dec. 1995.

No context found.

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE/ACMTransactions on Networking, 3:753--764, 1995.

No context found.

I. Katzela and M. Schwartz. Schemes for fault identification in communications networks. IEEE/ACM Transactions on Networking 3(6):753--764, December 1995.

No context found.

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE/ACM Transactions on Networking, 3(6):753--764, December 1995.

No context found.

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE/ACM Transactions on Networking, 3(6):753--764, 1995.

No context found.

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE/ACM Transactions on Networking, 3(6):753--764, December 1995.

No context found.

I. Katzela and M. Schwartz. Schemes for fault identification in communication networks. IEEE/ACM Transactions on Networking, 3(6):753--764, December 1995.

First 50 documents

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