作者: Giulia De Santis , Abdelkader Lahmadi , Jerome Francois , Olivier Festor
DOI: 10.1109/NTMS.2016.7792461
关键词: Darknet 、 Artificial intelligence 、 Hidden Markov model 、 Time windows 、 Markov process 、 Data mining 、 Computer science 、 Poisson distribution 、 Machine learning 、 Scale (map) 、 Basis (linear algebra)
摘要: We propose a methodology based on Hidden Markov Models (HMMs) to model scanning activities monitored by darknet. The HMMs of are built the basis number scanned IP addresses within time window and fitted using mixtures Poisson distributions. Our is applied real data traces collected from darknet generated two large scale scanners, ZMap Shodan. demonstrated that models able characterize their activities.