Search Results

The Link flooding Attack (LFA) Arises As A New Classof Distributed Denial Of Service (DDoS) Attacks In Recent Years.By Aggregating Low-rate Protocol-conforming Traffic To Congestselected Links, LFAs Can Degrade Connectivity Or Saturate Targetservers Indirectly. Due To Fast Proliferation Of Insecure Internetof Things (IOT) Devices, The Deployment Of Botnets Is Gettingeasier, Which Dramatically Increases The Risk Of LFAs. Since Theattacking Traffic May Not Reach The Victims Directly And Is Usuallylegitimate, LFAs Are Extremely Difficult To Detect And Defendby Traditional Methods. In This Work, We Model The Interactionbetween LFA Attackers And Defenders As A Two-person Extensiveform Bayesian Game With Incomplete Information. By Using Actionabstraction And The Divide And Conquer Method, We Analyzethe Nash Equilibrium On Each Link, Which Reveals The Rationalbehavior Of Attackers And The Optimal Strategy Of Defenders.Furthermore, We Concretely Expound How To Adopt Local Optimalstrategies In The Internet-wide Scenario. Experimental Resultsshow The Effectiveness And Robustness Of Our Proposed Decision-making Method In Explicit LFA Defending Scenarios.