As users of the Internet transition from an “always-connected” to an “information-sharing” society, it is imperative to address information retrieval for efficiency and scalability. Recent advances in the Internet to pivot from host-centric to information-centric architecture is a clean slate approach to treat information as a first-class addressable entity. This direction, known as Information-Centric Network (ICN), addresses efficiency, scalability, and robustness required to distribute and manipulate the ever-increasing information generated by applications and connected devices on the Internet.

This thesis introduces Directory Assisted Routing of Content in Information-centric networks (DARCI), Multi-Instantiated contents with Directory Assisted Routing (MIDAR) in Information-centric networks and Adaptive Cache Enabled Directory (ACED) assisted routing in Information-centric networks.

DARCI is a new architecture for information addressing, discovery, and routing where information names are mapped to distributed directories in a trie with prefix labels using a space-filling curve (e.g. Hilbert curve). DARCI combines information name and mapping to intervals and prefix labels to create a stateless, scalable, reliable, and efficient name resolution and routing of information between publishers and subscribers. DARCI achieves lookup speed in the order of magnitude higher than the non-deterministic Longest Prefix Match lookup on variable-length content name string by transforming content names to locality preserving codes of fixed length integers. MIDAR adapts the DARCI architecture to enhance further and scale information name resolution and routing for multiple instances of information. Routers and directories use the compact routing in intervals and prefix labels to analytically decide on the next direction for the request and response by combining knowledge of intervals and prefix reachability at a given node. ACED furthers DARCI and MIDAR to achieve an unconventionally efficient caching methodology where caches opt to become authoritative owners of selective information temporarily. Such a cache node registers the information name with the directory trie. Subsequent information name resolution receives the cache node as one of the publishers. Subscribers and intermediate nodes use this information to reach the closest copy of the requested content.