Dissertation > Excellent graduate degree dissertation topics show

Invulnerability Optimization and Evaluation Techniques of Complex Network

Author: LiuZuoNi
Tutor: ChenShanZhi
School: Beijing University of Posts and Telecommunications
Course: Computer Science and Technology
Keywords: Complex networks Invulnerability Model optimization Network survivability assessment
CLC: O157.5
Type: PhD thesis
Year: 2011
Downloads: 304
Quote: 0
Read: Download Dissertation

Abstract


Self-organization, self-similar attractor, small world, scale-free part or all of the nature of network is called a complex network. The nature and the social system complexity can be attributed to an interleaving network complexity, through these complex networks, the various components of the system between the dominant nonlinear effects. In recent years, the study of complex networks came into being, it is just the rise of a research direction in the course of the study of complex networks, scientists have found complex network of power-law distribution characteristics of, ie complex network node value K meet the power law relationship with respect to its probability P (K), and the exponent in the range of greater than 2 is less than 3. People from the network of this nature a special name: the scale-free network. The scale-free network lacks a feature value (or average value), fluctuations in the value of node degree range. The complex network of power-law features make the network can cause total paralysis that deliberate attacks show a strong vulnerability to attack the network only need to destroy a small number of key nodes; On the other hand, complex network network appear random attacks has strong robustness, which is a complex network robust and fragile characteristics. Therefore, a growing number of research institutions regard the complex network survivability as a hot issue. Complex network invulnerability sex the following aspects of the problem to be solved: 1) the understanding of the nature of the topological characteristics of the network rendering is still not deep enough: the network invulnerability study must have a better understanding of the topological characteristics of the network , that understanding of the causes of the network characteristics of the power-law, to solve the existing problems in the complex network from the internal laws of the network development. 2) optimization of the network survivability is limited to graph theory and mathematical methods, and not from the internal laws of the network development, ignoring the survivability as a role in affecting the cause of the network. 3) failure model established for the existing network cascading failure phenomenon, can not be considered from the attributes of the node itself; 4) existing network survivability assessment criteria is too complex, the lack of a simple and effective survivability assessment technology. This paper focuses on the question, to carry out the following research: 1) to optimize the trade-off / put-HOT (Highly Optimized Tolerance / Tradeoff) theory, and optimized from the system development point of view of the complex network characteristics of the power-law causes. The characteristics of the power-law is a common phenomenon in complex networks, there are a variety of explanations for the causes of this paper by a simple forest fire model HOT theory in the process of system optimization How to get the output of the power-law distribution characteristics, and the HOT Theory in system optimization and the establishment of the network model summary of the application. 2) put forward the based Hot theory network survivability dynamic evolution model: HOT theory is the relationship between the robustness of the system and the vulnerability of a hot research on complex systems, where the system must be its environment is robust in the face of uncertainties, the complexity of the system, however, makes this the robustness of the uncertainties of the outside world with a certain selectivity. HOT theory, considering the evolution of nodes in the network, the network of the study a number of factors affect the node joins the network, and survivability as a new node added to the network in a measure, and ultimately establish with anti- survivability of network dynamic evolution model. Theoretical analysis shows that the model is consistent with the reality of the Internet topology characteristics, another simulation results show that the model can effectively resist attacks in the network. 3) a node importance based on the Analytic Hierarchy Process - AHP (Analytic Hierarchy Process) network cascading failure model: cascade network failure phenomenon based node importance AHP network level Parallel Failure Model. Which the degree of importance of IMP node by node K, shortest path node S, and node neighbors shortest path NE Three factors determine the proportion of a variety of factors in the decision of the node importance AHP Analytic Hierarchy method calculated. Simulation analysis of the ER and BA networks in this cascading failure model changes in network efficiency by different types of attacks to prove the effectiveness of this model; another simulation analysis also examined the network when a variety of factors, the proportion of change the efficiency changes after the attacks, and ultimately show that the model can reduce network efficiency, decrease network under attack. 4) a node importance of entropy-based network survivability assessment techniques: scale-free characteristics of the network is actually a non-homogeneous characteristics, most of the nodes in the scale-free network nodes have only a small amount of connection, while a few nodes but has a very high degree, the this node degrees lead to uneven distribution of the characteristics of the network is difficult to effectively resist the intentional attacks, network to a certain extent reflect the network survivability. This paper proposes a network node importance entropy-based anti survivability evaluation algorithm, in which the importance of the nodes is determined by its betweenness, a good measure of the importance of the node in the network. After the attacks of different sizes under network entropy calculation can verify that this method can effectively evaluate network survivability.

Related Dissertations

  1. The Effect of Emotion on the Invulnerability of Representations in Visual Working Memory,B842.3
  2. Feature based on complex network propagation characteristics of SNS social networking sites,G206
  3. FEICHENG blending coal distribution center model,F259.2;F224
  4. Research on Application of Visualization of Analysis and Calculation in Power System,TM769
  5. Cone Crusher structural performance parameter optimization,TD451
  6. Complex network reliability evaluation research,O157.5
  7. Software-based research affecting network software metrics,TP311.52
  8. Attack directed repair complex network invulnerability Strategy,O157.5
  9. Transport Network Evaluation Based on Vehicle Travel Characteristics,U491.13
  10. Modeling for the Vehicle Running Environment of Road Traffic Accident,U491
  11. Research on Modeling Product Network and Characteristics for Enterprise Cluster,F273.2;F224
  12. The Study on Hydraulic Balancing Optimization of Fluid Pipeline Network,TV134
  13. Ruin Probability for Risk Model under Stochastic Environment & Stochastic Process in Complex Networks,F840
  14. Backbone of public transport network mode,U491.17
  15. Complexity of Public Transport Network in Tianjin,U491.17
  16. The Research on Functional Architectural Model of Forestry Application System Based on Complex Networks,S757.1
  17. Extraction of Coptis Chinensis Franch Alkaloids and Hypoglycemic Activity Research,R285
  18. Research of Interest Dissemination on Virtual Community,TP393.094
  19. Based on Virtools technology, three-dimensional roaming the campus network system design and implementation,TP393.18
  20. Research of Scientific Papers Screening Method Based on Complex Network of Different Mechanisms,G311
  21. A Study on the Credit Risk and the Optimization of Structure Based on VaR,F832.4

CLC: > Mathematical sciences and chemical > Mathematics > Algebra,number theory, portfolio theory > Combinatorics ( combinatorics ) > Graph Theory
© 2012 www.DissertationTopic.Net  Mobile