ADMBIFA: Accurate Detection and Mitigation of Blended Interest Flooding Attacks in NDNs

Abstract

The Named Data Network (NDN) is a future network architecture to transit conventional host-centric networks into data-centric networks. A routing attack, represented by Interest Flooding Attack (IFA), is a major security concern in NDNs. IFAs seriously harm NDNs by sending excessive fallacious Interest packets to overwhelm the routers to serve legitimate users. It is easy to launch, but difficult to defend, especially when legitimate Interest packets are blended with the IFAs as bIFAs. In this paper, we propose an accurate approach to detect and mitigate IFAs and bIFAs, named Accurate Detection and Mitigation of Blended Interest Flooding Attacks (ADMBIFA). It utilizes fuzzy logic to detect IFAs and bIFAs. And then it identifies malicious Interest prefixes with K-means. Finally, it restrains attackers to mitigate the damage to the networks. The simulation results show that the ADMBIFA is very sensitive to detect attacks with an ability to mitigate IFAs and bIFAs. And it can effectively reduce the negative impacts on NDNs.