Memory management optimization for content routers in DONA
Résumé
Nowadays, content retrieval is marking the Internet usage. User communications are no longer tied up to host interconnection. Information Centric Networking (ICN) models are proposed to cope with these changes. The new paradigm redesigns the Internet architecture to bring out content to the first level. Over the last decade, many key projects have proposed a large solution spectrum to rebuilt networking primitives focused on the content. One important and direct challenge of this shift is the large amount of routing states due to identifying contents rather than hosts.
In this paper, we focus especially on DONA, one of the first ICN architecture, and analyse the required memory space to store routing states. Our study shows that today’s technologies
are not able to satisfy the content routing needs. Thus, we propose an enhancement of DONA called BADONA to deal with this problem. It uses a Bloom filter to drastically reduce the
usage of the memory space.