Community-based Malware Immunization Strategy

Abstract

In recent years, security threats have occurred frequently, especially the large-scale spread of malware has brought huge losses to individuals and the country as a whole. This paper considers the nodes of the network have the feature of forming a community. A community-based malware immunization strategy is proposed according to the characteristics that information spread quickly within the community and spread slowly between communities. A concrete community detection method is proposed based on the K-shell and the label propagation algorithm, as well as key node selection algorithm is proposed in order to control the outbreak of malicious code with the fastest speed and at the lowest cost. The experiments demonstrate that the proposed approach can achieve maximum immunity with minimal cost comparing to classical immunization strategies.