Maximizing network interruption: an intruder's perspective

Abstract

With the colossal growth of Wireless Sensor Networks (WSNs) in different applications starting from home automation to different military affairs, the pressure on ensuring security in such a network is paramount. Considering different security challenges, it is really a tough job to develop a secured sensor network system. Moreover, as technology is getting popular, the intruders also planning new ideas to break the security of a system, to harm the network and make the system quality down, thus, take the control of the network to corrupt it anyhow. In this thesis, we deal with the security breaking cost of a network, how to make it higher for an intruder with the goal of making it tougher for him to break the system security. From an intruder’s point of view, we study, in case of multi-hop wireless network, how easy it is for an intruder to completely jam all the sensor nodes in the network. Here, we call the problem as Maximizing Network Interruption Problem (MIP) where we assume that the intruder has the complete picture of the entire network and therefore is able to strategically place interrupting or jamming nodes at suitable locations of its own desire to compromise the network operation by jamming the wireless channel. The goal is to tactically place the malicious or jamming nodes in the sensor network, so that all the sensor nodes in the network are jammed or corrupted and number of jamming or malicious node required is minimum or near minimum. In this thesis, we proposed four approximation algorithms to solve the MIP problem with necessary figures and descriptions. Experimental results have been shown for better understanding of each scenario and finally comparative analysis has been added to evaluate the four approximations and their different branch of applications.