Towards attack-agnostic defenses

摘要: Internet attackers control hundreds of thousands to perhaps millions computers, which they can use for a variety different attacks. Common attacks include spam delivery, phishing, and DDoS. The current research community focus is on defenses each specific attack type compromised hosts may launch. However, attack-specific approaches almost always have two fundamental drawbacks: do not address the root problem that an army hosts, provide right incentives users properly secure their machines. As result, be defeated by new attacks, even those only slightly from old attacks. We argue researchers should also attackagnostic whose effectiveness does depend particular type. We initiate this line investigating design space attack-agnostic defenses, then detailing extreme points within space: InternetWatch List Reputation System.