Stateless Forwarding in ICN with Dynamic Non-Bloom Filters

In this paper, we propose a novel packet forwarding architecture for information centric networking (ICN) towards achieving resource efficient and scalable content delivery using ICN principles. Proposed solution addresses one of the main challenges for ICN, by focusing on the limitations of stateful forwarding; which is utilized by content-centric networking (CCN) based ICN architectures at the ICN-enabled content routers to provide breadcrumbs for content objects to follow by creating forwarding states for their requests along the path towards the content source. Due to increased storage requirements associated with stateful forwarding, it can potentially lead to inefficient packet forwarding to prevent routers from forwarding at line-rate speeds. Existing solutions focused on using Bloom filter (BF) based designs to achieve stateless forwarding, however, at the cost of false positives. Unlike these earlier works, our solution addresses this problem by proposing a dynamic BF-less scheme to not only address the stateful forwarding objectives but also to limit problems associated with the existing stateless forwarding designs that rely on BFs. Through a simulation-based study, we observed significant 5-20 times improvement in transmission overhead with the proposed solution when compared to BF-based stateless forwarding, while offering a more processing-friendly design.