Investigating Impact of Fake-Type Content Poisoning Attack on NDN

Information-centric networking (ICN) has been widely studied as a new network architecture for efficient content delivery. However, the threat of content poisoning attack (CPA) has been pointed out. In the CPA, a malicious user degrades the cache hit ratio of router caches by inserting fake contents into router caches. To investigate effective countermeasures against the CPA, we need to clarify the impact of the CPA on reducing the cache hit ratio. However, existing studies assumed limited CPA schemes, and they were based on small network topologies without considering caching methods and locations of fake contents and bots. In this paper, we focus on the original-fake type CPA which is the easiest CPA for attackers and most difficult to be detected, and we investigate the potential threat of the original-fake CPA for various caching methods on large-scale networks when optimally assigning locations of fake contents and bots.