Overview

The goal of this project is to analyze SCAN's performance and operation mathematically, and evaluate it. A recent measurement study shows that most of the Internet traffic is governed by content retrieval applications. However, the current Internet was designed based on the host-to-host communication paradigm, and hence inefficient content deliveries often happen. To solve the inefficiency, information-centric networking (ICN) introduces a new networking paradigm that is consistent with the current Internet usage pattern. To achieve the goal of efficient content networking, this project develops network software based on the technical concept named, Content-Centric Networking (CCN). CCN is an evolving distribution architecture envisioned by PARC that enables efficient and secure distribution of content than is currently feasible.

To exploit the cached contents, prior researches on ICN usually rely on an opportunistic cache-hit (happen-to-meet), which can utilize contents cached only in routers on the path towards the publishers. However, considering the portion of in-network storages on the path and the spatial diversity of content requests, the opportunistic cache-hit limits the potential advantages of network-wide in-network storages. Addressing the above issues, we propose a scalable content discovery scheme for ICN, dubbed SCAN. In SCAN, Interests are spreaded to off-path content routers as well as an original content publisher.

Generally, off-path routing can utilize closer caches. However, to do this, prior data exchanges(such as cached content information) are needed, and this may introduce some side effects in terms of network traffic.

Administrative Data

  • Overall Period: May. 1 2013 ~ April. 30. 2014
  • Sponsor: Cisco

Members

(as of March 2012)

Keywords

(as of March 2012)

  • ICN, SCAN, cache, off-path, modeling