Font Size: a A A

Research And Implementation Of Fault Management Technology

Posted on:2012-02-01Degree:MasterType:Thesis
Country:ChinaCandidate:J YanFull Text:PDF
GTID:2218330362951916Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Communication network fault management technology is the core of network management system is to maintain efficient operation of the key factors in the network, and fault location detection of network fault management as one of the key functions of its technology, good or bad algorithm will determine the automation of the entire network success or failure. This article is the current network system for different levels of alarm because of missing and false alarm situation, resulting in reduced accuracy of fault detection, rapid diagnosis of the problem can not be put forward corresponding solutions.Firstly, for the TCP/IP-protocol network layer management principles sampling, which makes detection of the respective layers of the network entities are the characteristics of symptoms, a multi-functional service-oriented model of hierarchical fault diagnosis of MFS-LFDM. The model is based fault propagation model bipartite graph algorithms, based on a top-down approach to network fault diagnosis fault node identifier FNDI to decompose, and then delivered to the business service layer, network communication layer and application clients One layer of internal processing, has greatly improved the efficiency of fault analysis for rapid and accurate fault location later provides the basis.Secondly, for the loss of network systems and false alarms in the alarm will cause the accuracy of fault location detection reduced the problem of increased latency, a three-node architecture based on the fault location algorithm TNSFLA. Construction of the algorithm include the failure of a node, the node and alarm diagnosis, including three-tier system architecture nodes TNS, using the framework of fault location model, the mathematical formula of the problem, and ultimately it comes down to strike the fault diagnosis of the biggest queues confidence issues. Algorithm implementation process includes two stages, first create a network system can explain the set of nodes receiving the warning symptoms of the corresponding set of fault diagnosis queue, the second stage is the diagnosis of a queue created in the various elements, all in accordance with formula failure occurred in a subset of elements of confidence, confidence to choose the largest element of the fault diagnosis as the final result.Finally, the simulation of fault detection rate of the algorithm, the failure rate of false detection, fault location time performance was tested, the results showed that when the network is lost and false alarms exist in the case of failure, the algorithm in the fault detection rate, false Detection rate and the positioning time, shows that the great superiority.
Keywords/Search Tags:Network Fault Management, Fault Location, Hierarchical Diagnosis Model
PDF Full Text Request
Related items