作者: Anton ROMANOV , Eiji OKAMOTO
DOI: 10.1587/TRANSCOM.E93.B.1411
关键词:
摘要: With the increasing demand for services provided by communication networks, quality and reliability of such as well confidentiality data transfer are becoming ones highest concerns. At same time, because growing hacker's activities, content its continuous delivery strongly depend on integrity transmission availability infrastructure, thus information security a given IT landscape. But, amount resources allocated to provide (like staff, technical countermeasures etc.) must be reasonable from economic point view. This fact, in turn, leads need employ forecasting technique order make planning budget short-term potential bottlenecks. In this paper we present an approach wide class related incidents (ISRI) — unambiguously detectable ISRI. is based different auto regression models which widely used financial time series analysis but can not directly applied ISRI due specifics security. We investigate address proposing rules (special conditions) collection storage series, adherence improves subject field. application our one type spam messages which, if mitigated properly, could create additional load infrastructure consume significant amounts network capacity. Finally evaluate simulation actual measurement.