Avoiding Notorious Content Sources: A Content-Poisoning Attack Mitigation Approach

Abstract

Named Data Networking (NDN) has emerged as a promising Future Internet architecture. NDN provisions security by design and guarantees that data packets are immutable and authentic. Nevertheless, its inherent in-network caching feature has opened the door to new types of security attacks. One such critical security issue in NDN is content poisoning attacks. In content poisoning, the attacker aims at injecting poisonous (i.e., fake or invalid) content in the network caches. In this paper, we propose a reputation-based content poisoning mitigation model, which assists both the access and the core network nodes in identifying the sources from which poisonous content is originated, and subsequently, limiting the Interest flow towards those notorious sources as well as in avoiding caching poisonous content.