The Research And Realization Of Effective And Adaptive Active Probing Algorithm

As one of the core functions of network management, fault management can detect and locate faults in the network quickly and accurately. It can reduce the loss caused by network faults, assure the availability and reliability of network and prevent the occurrence of faults to a certain extent. The current fault management systems locate faults through alarm monitoring and correlation. It requires the target system has the ability to report its inner message and it's impossible to detect and locate faults before user senses them. Sometimes, it even can't locate faults quickly because of delay and loss of the alarms. All of these make the fault management system can not satisfy requirements of complex network. Fault localization based on active probing is an active method to get network information and a probe can get the real-time element status on the probe path. The activeness and adaptiveness of active probing can make up for the disadvantages of alarms. And it has great application potential in fault management.This paper studies the basic method of fault localization using active probing, analyses the key problems of active probing, and provides some effective solutions to these problems. This paper also shows how to use active probing in fault management through a fault management system case. The contents of this paper includes:1) Study and summarize the general methods and key problems of using active probing in fault management, including process and algorithm.2) Analyze and improve the probe station selection algorithm and propose an effective algorithm based on vector span.3) Analyze and improve probe selection algorithm in fault detection to promote the execution efficiency of algorithm.4) Analyze and improve probe selection algorithm in fault localization and propose an effective and adaptive algorithm; Analyze how to improve the anti-noise ability of probe and introduce the average Hamming distance of the probe dependency matrix as the standard to measure the anti-noise abilities of probe sets.5) Verify the effectiveness of the algorithms through simulation experiments and analyze the experiments'results.6) Show the application of active probing in fault management.