Characterization of Evolving Networks for Cybersecurity

The process of network evolution presents an interesting problem that describes shifts in the behavior of a network structure. Given the widespread use of computer networks creates a need to drill down and analyze the behavior at the various levels to get a tangible perspective on shifts that may take place in a computer network. However, with cyber attacks becoming increasingly sophisticated, one of the key challenges is knowing whether there is even an attack on the network in the first place. This is mainly due to the overwhelming size and dynamism of evolving network structures which poses a big data problem. This chapter discusses graph theory concepts to model network behavior and then utilizing analytics to understand the dynamics of the network. For example techniques such as graph sampling play an important role in identifying potential cyber threats that may not be detected on a larger scale. Additionally, determining micro and macro level characteristics that are associated to key network features such as node centrality, and fundamental network properties such as densification and diameter respectively are critical in characterizing network behavior overtime that may be the result of a cyber threat. Therefore, in order to define and understand the vulnerabilities associated to the network, one must have an understanding of the overall structure and nature of communication patterns within the network as well as the potential points of vulnerability.