Simulation based analysis of spreading dynamics of malware in wireless sensor networks

Abstract

We analyze spreading dynamics of malware (malicious software) that can replicate itself on other nodes in a network of wireless sensing devices. Epidemic models have previously been employed by researchers to investigate the spreading dynamics of malware in the Internet. Most studies have generally ignored spatial deployment characteristics of nodes which are critical for a realistic analysis in wireless sensor networks. We consider random and localized attack strategies by malware that may be introduced by an adversary into a network with either completely random or clustered sensor deployment patterns. Results demonstrate that sensor networks face serious security threats, particularly in cases where sensors are deployed in clusters and when malware can carry out attacks to random targets. We recommend possible defense measures based on simple topology control that might slow down the spread of malware across the network and allow necessary measures to be taken for containing the spread when malware employs a localized attack strategy. Presented analysis has a significant impact on network security and defense industry since it provides an insight into potential threats to information and surveillance infrastructure consisting of wireless sensing devices. © 2007 IEEE.