作者: Elias Bou-Harb , Mourad Debbabi , Chadi Assi
DOI: 10.1016/J.COSE.2014.02.005
关键词:
摘要: Motivated by recent cyber attacks that were facilitated through probing, limited security intelligence and the lack of accuracy is provided scanning detection systems, this paper presents a new approach to fingerprint probing activity. It investigates whether perceived traffic refers activities which exact technique being employed perform probing. Further, work strives examine dimensions infer ‘machinery’ scan; random or follows certain predefined pattern; strategy employed; activity generated from software tool worm/bot. The leverages number statistical techniques, probabilistic distribution methods observations in an attempt understand analyze activities. To prevent evasion, formulates matter as change point problem yielded motivating results. Evaluations performed using 55 GB real darknet shows extracted inferences exhibit promising can generate significant insights could be used for mitigation purposes.