C. Hood and C. Ji. Proactive network fault detection. In Proc. IEEE INFOCOM '97, Kobe, Japan, April 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....focused 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 ....

C. Hood and C. Ji, "Proactive network fault detection," in Proceedings of IEEE INFOCOM '97, Kobe, Japan, April 1997.

....and correct model is difficult to be achieve, expert systems and neural networks can diagnose faults based on the typical symptom based associations. The drawbacks are that it is time consuming and that it does not avoid all disadvantages of all approaches included in the model. Hood and Ji [HJ97] also propose a hybrid approach combining a probabilistic method that provides 12 a Bayesian model of the network, with a proactive learning system to set parameters. Another hybrid approach has been proposed by Gardner and Harle [GH97] They propose to combine a special type of ANN called ....

C.S. Hood and C. Ji. Proactive Network-Fault Detection. IEEE Transactions on Reliability, 46(3), September 1997.

....with a variety of approaches competing for acceptance. An e ective solution to the problem of event correlation has not been popularly accepted. Many research e orts that claim to be innovative appear to be unique solu 13 tions that would be dicult to adapt widely; portability is not emphasized. [15], 29] 1] 26] 9] Proprietary products that claim to use event correlation successfully are large, costly, integrated software modules. 23] 12] 21] 19] The corporate website typically publishes white papers that present their company s engineering approach to correlation in enough ....

....other parts of the system. Though dicult to implement in practice, the research shows promise of being able discover network problems before they occur. This approach has had limited 17 application in research; no commercial products are known to use this technique, but is being seriously studied. [15], 43] 41] Codebooks The coding based approach attempts to create noise insensitive correlation. The relationships between known event sequences are characterized into a correlation matrix, essentially a binary matrix that maps events to alarms. That is, a matrix of ones and zeros, where ....

C. S. Hood and C. Ji. Proactive network fault detection. Proceedings of INFOCOM 1997.

.... that faults can be detected by statistical deviations from regularly observed behavior [11] Brutlag which applies thresholds to time series models to detect aberrant network behavior [12] and Hood and Ji who describe an adaptive monitoring system which is able to detect unknown or unseen faults [13]. Most of this work focuses on how to detect accurately deviations from normal behavior, whereas our work is focused on analyzing and characterizing statistically specific types of anomalous behavior. Many papers have been written on detection of nefarious behavior such as denial of service (DoS) ....

C. Hood and C. Ji, "Proactive network fault detection," in Proceedings of IEEE INFOCOM '97, Kobe, Japan, April 1997.

....di#culties, we are strongly motivated to develop predictive models of networked systems, since having such models will enable a new generation of proactive management technologies. There has been considerable works in the area of fault detection [6, 7] There are some e#orts in proactive detection[5, 8]. Our work can be viewed as a sequel to [8] This paper studies the number of hypertext protocol (HTTP) operations per second. The data are detrended using a model of daily tra#c patterns that consider time of day, day of week, and month, and an autoregressive model is used to characterize the ....

C.S.Hood, C. Ji, "Proactive Network Fault Detection", Proceedings of INFOCOM, Kobe, Japan, 1997.

....been used in protocols for reliable multicast to determine tree topologies [17, 18] In these schemes, receivers use multicast traceroutes to discovers the nearest peer on the other side of a bottleneck link. Of relevance to our work are techniques for proactive fault isolation. One such approach [14] learns normal behavior of the network by developing estimates for router state, then detects and ags deviations from these estimates. Such a system can detect abnormal behavior before a fault actually occurs. Similar learning techniques can be employed in our algorithms as well. Some of our ....

Cynthia S. Hood and Chuanyi Ji. Proactive Network Fault Detection. In Proceedings of the IEEE INFOCOM, Kobe, Japan, April 1997.

....tools are initiated to monitor the same session. This is because of the uncontrolled multicast traceroute tra c that can be potentially generated by the various MHealth tools monitoring the same session. Of relevance to our work are techniques for proactive fault isolation. One such approach [24] learns normal behavior of the network by developing estimates for router state, then detects and ags deviations from these estimates. Such a system can detect abnormal behavior before a fault actually occurs. Similar learning techniques can be employed in our session watchers as well. 5 ....

Cynthia S. Hood and Chuanyi Ji. Proactive network fault detection. In Proceedings of the IEEE INFOCOM, Kobe, Japan, April 1997.

....been used in protocols for reliable multicast to determine tree topologies [16, 17] In these schemes, receivers use multicast traceroutes to discovers the nearest peer on the other side of a bottleneck link. Of relevance to our work are techniques for proactive fault isolation. One such approach [13] learns normal behavior of the network by developing estimates for router state, then detects and ags deviations from these estimates. Such a system can detect abnormal behavior before a fault actually occurs. Similar learning techniques can be employed in our algorithms as well. Some of our ....

Cynthia S. Hood and Chuanyi Ji. Proactive Network Fault Detection. In Proceedings of the IEEE INFOCOM, Kobe, Japan, April 1997.

....been used in protocols for reliable multicast to determine tree topologies [33, 34] In these schemes, receivers use multicast traceroutes to discovers the nearest peer on the other side of a bottleneck link. Of relevance to our work are techniques for proactive fault isolation. One such approach [24] learns normal behavior of the network by developing estimates for router state, then detects and ags deviations from these estimates. Such a system can detect abnormal behavior before a fault actually occurs. Similar learning techniques can be employed in our session watchers as well. 6 ....

Cynthia S. Hood and Chuanyi Ji. Proactive network fault detection. In Proceedings of the IEEE INFOCOM, Kobe, Japan, April 1997.

.... instrumentation [4, 9, 11] performance prediction [8, 10] and program steering [3, 13] Additionally, adaptive techniques have been explored for predictive signaling and control in cluster environments for performance management [12] and in highly distributed networks for use in fault management [5]. Our focus, however, is on soft faults. Specifically, we wish to understand the mechanisms behind network contributions to soft faults and to identify ways to signal or ultimately control such faults. 3. The Problem of Cluster Fault Management As high performance systems, clusters require ....

C. S. Hood and C. Ji. Proactive network-fault detection. IEEE Transactions on Reliability, 46(3):333--341, September 1997.

....scheme on the MIB variables [11] choice of an optimal set of variables, study of the statistical properties of the MIB variables, and the generalization capability of the algorithm to another node. The potential use of MIB variables in proactive fault management has been explored previously in [12][13] In the present work a new, simpler fault detection algorithm has been developed which takes advantage of the interrelationships between MIB variables through a simple combination scheme, generating time correlated node level alarms. An appropriate subset of MIB variables was chosen based on ....

C.S. Hood and C. Ji. Proactive network fault detection. Proceesdings of INFOCOM, 1997.

No context found.

C. Hood and C. Ji. Proactive network fault detection. In Proc. IEEE INFOCOM '97, Kobe, Japan, April 1997.

No context found.

C. Hood and C. Ji. Proactive Network Fault Detection. In IEEE INFOCOM, 1997.

No context found.

C. Hood and C. Ji. Proactive Network Fault Detection. In IEEE INFOCOM, 1997.

No context found.

C. Hood and C. Ji, "Proactive network fault detection," IEEE Trans. Reliability, vol. 46, no. 3, pp. 333--341, 1997.

No context found.

C. Hood and C. Ji. Proactive Network Fault Detection. In IEEE INFOCOM, 1997.

No context found.

C.S. Hood and C. Ji. Proactive network fault detection. In Proceedings of INFOCOM, 1997.

No context found.

C. Hood and C. Ji. Proactive network fault detection. In Proceedings of IEEE INFOCOM '97, April 1997.

No context found.

C. Hood and C. Ji. Proactive network fault detection. In Proc. IEEE INFOCOM '97, Kobe, Japan, April 1997.

No context found.

C. Hood and C. Ji. Proactive network fault detection. In Proc. IEEE INFOCOM '97, Kobe, Japan, April 1997.

No context found.

C. Hood and C. Ji. Proactive network fault detection. In Proc. IEEE INFOCOM '97, Kobe, Japan, April 1997.

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