Title
Detection of collusive interest flooding attacks in named data networking using wavelet analysis.
Abstract
Named Data Networking (NDN) is a clean state Internet paradigm that considers some security primitives in its original design, and one of the promising candidates for the future Internet architecture. However, it may suffer from the specific threats such as Interest Flooding Attacks (IFA). Most researches are focused on the defense against fake IFA, where the PIT resources in victim routers are exhausted by the flooding forged Interest packets requiring the non-existent contents. However, these existing mitigation mechanisms cannot defend the Collusive Interest Flooding Attacks (CIFA), which is disregarded by most related researches. In CIFA, the adversaries issue Interest packets that can be satisfied by a collusive server. In order to exhaust the PIT resources on the victim routers without being detected, the collusive server responds the received Interest packets with corresponding Data packets just before expiration of the corresponding PIT entries. In addition, the low rate intermittent characteristic of CIFA makes it difficult to be distinguished from the legitimate background traffic. Thus, the CIFA is more concealment and deception than the previous IFA. By analyzing the traffic model of CIFA, we discover that the CIFA is mostly concentrated in the low frequency bands. Based on this phenomenon, we extract the signal in attack sub-band using wavelet analysis technique and then propose our CIFA detection algorithm. We implement the CIFA model and evaluate our proposal in the ndnSIM simulator, and results confirm the feasibility and effectiveness against the CIFA.
Year
Venue
Keywords
2017
IEEE Military Communications Conference
Named Data Networking,Interest Flooding Attacks,Wavelet Analysis
Field
DocType
ISSN
Deception,Computer science,Network packet,Server,Computer network,Feature extraction,Flooding (psychology),The Internet,Wavelet transform,Wavelet
Conference
2155-7578
Citations 
PageRank 
References 
2
0.40
9
Authors
5
Name
Order
Citations
PageRank
Yonghui Xin162.81
Yang Li220815.35
Wei Wang392.63
Weiyuan Li431.08
Xin Chen561.48