Network Under Joint Node And Link Attacks: Vulnerability Assessment Methods And Analysis

ABSTRACT

Critical infrastructures such as communication networks, electrical grids, and transportation systems are highly vulnerable to natural disasters and malicious attacks. Even failures of few nodes or links may have a profound impact on large parts of the system. Traditionally, network vulnerability assessment methods separate the studies of node vulnerability and link vulnerability, and thus ignore joint node and link attack schemes that may cause grave damage to the network. To this end, we introduce a new assessment method, called β-disruptor, that unifies both link and node vulnerability assessment. The new assessment method is formulated as an optimization problem in which we aim to identify a minimum-cost set of mixed links and nodes whose removal would severely disrupt the network connectivity. We prove the NP-completeness of the problem and propose an O( ) bicriteria approximation algorithm for the disruptor problem. This new theoretical guarantee improves the best approximation results for both link and node vulnerability assessment in literature. We further enhance the proposed algorithm by embedding it into a special combination of simulated annealing and variable neighborhood search method. The results of our extensive simulation-based experiments on synthetic and real networks show the feasibility and efficiency of our proposed vulnerability assessment methods.

AIM

The main aim of this paper is the new assessment method is formulated as an optimization problem in which we aim to identify a minimum-cost set of mixed links and nodes whose removal would severely disrupt the network connectivity

SCOPE

The scope of this paper is the results of our extensive simulation-based experiments on synthetic and real networks show the feasibility and efficiency of our proposed vulnerability assessment methods.

 EXISTING SYSTEM

The first proposed the pair wise connectivity as an effective measurement and used mathematical programming to solve for exact solutions. and later defined the Critical Node/Edge problems, in which the main objective is to identify top k nodes/links that removal minimizes the pair wise connectivity in the residual network, and provided NP-completeness proofs and integer programming formulations. However, the runtime for exact solutions scale exponentially with the network size. Proved that the critical node detection problem (CNP) is also NP-complete. On trees for the total weighted pair wise connectivity metric. Proved that the CNP is polynomially solvable in trees and series-parallel graphs for the cases when the nodes have uniform costs and the objective is either minimizing the size of the largest component or maximizing the number of residual components.

 DISADVANTAGES

·      Even failures of few nodes or links may have a profound impact on large parts of the system

·      Traditionally, network vulnerability assessment methods separate the studies of node vulnerability and link vulnerability, and thus ignore joint node and link attack schemes that may cause grave damage to the network.

 PROPOSED SYSTEM

In this paper, study the effect of joint node and link attacks in terms of connectivity. We introduce a new problem, called β-disruptor,that finds a minimum-cost set of nodes and links whose removal degrades the pair wise connectivity to a great extent (a fraction). The β-disruptor problem aims to provide a more comprehensive assessment on network vulnerability. It generalizes both the β-vertex disruptor and the β-edge disruptor problems proposed in our previous work. To our best knowledge, this is the first work to address the effect of simultaneous attacks on both links and nodes on network pair wise connectivity.

 ADVANTAGES

1.  Finds a minimum-cost set of nodes and links whose removal degrades the pair wise connectivity to a great extent (a fraction ).
2. The β-disruptor problem aims to provide a more comprehensive assessment on network vulnerability

 System Configuration

Hardware Requirements

• Speed                  -    1.1 Ghz
• Processor              -    Pentium IV
• RAM                    -    512 MB (min)
• Hard Disk            -    40 GB
• Key Board                    -    Standard Windows Keyboard
• Mouse                  -    Two or Three Button Mouse
• Monitor                -     LCD/LED

 Software requirements

• Operating System              : Windows 7             
•  Front End                           : ASP.Net and C#
• Database                             : MSSQL
• Tool                                    : Microsoft Visual studio

References

Dinh, T.N,Thai, M.T.,“ Network Under Joint Node and Link Attacks: Vulnerability Assessment Methods and Analysis” IEEE/ACM Transactions on Networking,  Volume 23 ,  Issue 3 April 2014.