Information Assurance Considerations for a Fully Netted Force: Implementing Cranof for Strategic Intrusion Assessment for Cyber Command and Control
William C. Torrez, Donald Bamber, I. R. Goodman · 2001
We consider here a first step in applying CRANOF (Complexity Reducing Algorithm for Near Optimal Fusion - a new rigorously-based complexity- reducing algorithm that produces estimates of underconstrained probabilities - to the problem of determining when a computer network is under attack. Essentially, CRANOF treats this issue as a formal analogue of the pairwise track correlation or similarity problem, comparing the current cyber-state history with each of various alternative classes of cyber-state histories relative to various features or attributes measuring various degrees of normality / abnormality.