Efficient Content Verification in Named Data Networking

ABSTRACT In Named Data Networking, contents are retrieved from network caches as well as the content server by their name. This aspect arises severe security concerns on content integrity. Especially, if poisoned contents lie in the network cache, called content store(CS), interests would be served by the poisoned content rather than they propagate toward the content server. Consequently, users whose interests pass through the contaminated CS cannot access the valid content. In order to resolve the problem, every content is verified before they are inserted into the CS. However, this builtin verification mechanism is not practically feasible due to its huge computational overhead. In this paper, we address problems of content integrity in NDN in details, including how to violate content integrity. We also propose a practical solution that efficiently detects poisoned contents from the CS with minimum overhead. Since the proposed scheme aligns to the basic NDN architecture, it is a practical and effective solution