Predicting Global Trend of Cybersecurity on Continental Honeynets Using Vector Autoregression
Xing Ling, Yeonwoo Rho, Chee‐Wooi Ten · 2019
The deployment of honeynets from around world is intended to lure attackers into their networks and hence their footprint can be extracted and studied. This global trend can be correlated based on publicly available statistics. This paper proposes a statistical analysis to identify a geospatial and temporal patterns in the cyberattacks and use this knowledge to predict future attack trend. Using a publicly available honeypot data, this work aims to (i) incroporate long range dependence in the analysis of the number of cyber-attacks, which may be a result of spread of malware agents, (ii) propose a measure on how to determine whether or not to consider dependence structure between different honeypot hosts, and (iii) establish a modeling tool that would be intuitive in a honeynet, where honeypot hosts are closely connected and related. The proposed vector autoregression approach reveals the dependencies among honeypots.