Intrusion Detection Forecasting Using Time Series for Improving Cyber Defence

摘要: The strength of time series modeling is generally not used in almost all current intrusion detection and prevention systems. By having models, system administrators will be able to better plan resource allocation readiness defend against malicious activities. In this paper, we address the knowledge gap by investigating possible inclusion a statistical based that can seamlessly integrated into existing cyber defense system. Cyber-attack processes exhibit long range dependence order investigate such properties new class Generalized Autoregressive Moving Average (GARMA) used. GARMA (1, 1; 1, ±) model fitted cyber-attack data sets. Two different estimation methods are Point forecasts predict attack rate possibly hours ahead also has been done performance models discussed. investigation case-study confirm exploiting properties, it cyber-attacks (at least terms rate) with good accuracy. This kind forecasting capability would provide sufficient early-warning for defenders adjust their configurations or allocations.