Dissertation > Excellent graduate degree dissertation topics show

Research and Application of Burst Abnormality Detection Based on Network Traffic

Author: CaoMin
Tutor: ChengDongNian
School: PLA Information Engineering University
Course: Communication and Information System
Keywords: Network failure Fault Management Network Traffic Traffic Anomaly Detection Flow projections
CLC: TP393.06
Type: Master's thesis
Year: 2009
Downloads: 37
Quote: 0
Read: Download Dissertation


Network fault management is an important part of network management, fault management effectiveness and functional strength is directly related to the management network availability and reliability. Fault detection is the premise of network fault management, fault detection is an important means of network anomaly detection, network traffic data can be timely to reflect network fault conditions. Currently used is based on SNMP (Simple Network Management Protocol) network fault detection mechanism that Client / Server technology for centralized fault detection. If the mechanism Trap detection method, the over-reliance on the entire network, multi-node connection interrupt cause the whole network out of control; using the polling detection methods, the sudden failure detection noticeable lag. The mechanism can not detect large-scale, sudden network failure. Link traffic and routers to forward the amount of data the mutation and congestion result in traffic anomalies caused by large-scale, sudden network failure is defined as the sudden ruin strikes. For this type of event, proposes a new fault detection mechanism to detect nodes and links, and integrated to determine the cause of the fault and the division of the fault level. The test results failure to avoid around to provide a basis for rapid self-healing protocol and flow projections to determine the status of the network after the failure to avoid around. This article consists of the following aspects: (1) node failure detection, an adaptive threshold residual traffic anomaly detection algorithm. The algorithm is in the study time sequence detection algorithm and the residual ratio on the basis of the detection algorithm advantages and disadvantages, to achieve precise characterization of network behavior characteristics, and decreases the fixed threshold value of the false alarm and missed alarm object, the residual ratio detection algorithm fixed threshold to improve the adaptive threshold, the network traffic anomaly detection. Offline threshold calculated at the same time, to improve the detection speed. The simulation results demonstrate the accuracy and performance of the algorithm; (2) for the link fault detection, mutation detection algorithm research on the basis of the existing flow, mutation detection algorithm based on the flow of aggregate functions. The algorithm is based on an improved histogram technology, combined with aggregate functions query mutation detection of network traffic. The advantages of using histogram compression technology to store data, combined with aggregate functions adjacent isometric window traffic anomaly judgment. The detection accuracy and speed of the algorithm by simulation analysis; (3) the failure to avoid around the network status monitoring for rapid self-healing protocol proposed prediction algorithm based on Kalman filter network traffic. The algorithm by the basic principles of the Kalman filter with the actual situation of the network combined network modeling. The purpose of the algorithm is to verify potential energy after adjusting the network is working properly, whether it can maintain normal communication. To verify the the algorithm validity and correctness by simulation; (4) for the existing fault management system is difficult to meet the demand of topics, design a strike event management system based on abnormal network traffic burst destroyed. Application of the system in the network node router detection network traffic to detect network failures, the full three-dimensional protection network. Network system test results to determine cause of the fault, the fault classification and orientation notices, and ultimately through the fast recovery protocol adjustment mechanism to avoid network failure, and prediction algorithm to evaluate the adjusted network. The design system visualization user interface, combined with the off-line statistical methods, given the link-state index and network dependency graph, to guide subsequent fault detection and for future work based on the theoretical analysis.

Related Dissertations

  1. Campus Network Management Flow Analysis Technology Research and Implementation,TP393.06
  2. Traffic Prediction Model Based on Wavelet and Markov,TP393.06
  3. Research on Frame Traffic Control, Prediction and Energy Efficiency in Wireless Local Area Networks,TN925.93
  4. Study on the Energy Saving Technologies Based on Traffic Prediction in Cellular Net-works,TN929.53
  5. The Research and Implementation of the Fault Management System for Triple Play,TP315
  6. Based on non- parametric statistical characteristic quantities Gaussian kernel network traffic anomaly detection method,TP393.07
  7. The Research and Implement on the Application Protocol for Intenet of Things (IoT) Services,TN929.5
  8. Based on magnetoresistive sensors and ZigBee network traffic monitoring system ITS,TP277
  9. Vehicle self-organizing network connectivity study,TN929.5
  10. Research of Prediction Technology Based on Network Traffic Data Character Analysis,TP393.06
  11. Short-term Urban Traffic Forecasting Based on Multi-kernel SVM Model,U491.14
  12. Study on Technology of Emergency Traffic Organization for Regional Road Network under Sudden Events,U491
  13. Integrated Management and Implementation of SNMP DoS attack defense system,TP393.08
  14. Research on the Stream Data Mining in Network Traffic Analysis,TP393.06
  15. The Research of Freeway Network Traffic Regulation Based on Expense Function,F224;U491
  16. Improvement of Local Support Vector Machine and Application in the Network Traffic Forecasting,TP393.06
  17. Design and Implementation of Alarm Data Collection Subsystem of Mobile Telecom Network Management System,TP274.2
  18. A Study on Fault Diagnosis in Wireless Sensor Networks,TN929.5
  19. A Research of P2P Traffic Identification Based on Characteristic Process,TP393.02
  20. A Study and Application of Network Traffic Monitoring Technology Based on Campus Network,TP393.06
  21. Research on Sampling Algorithm for Traffic Measurement Based on Fairness Mechanism,TP393.06

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Computer network > General issues > Computer networks, test , run
© 2012 www.DissertationTopic.Net  Mobile