ChunkCast: An Anycast Service for Large Content Distribution
Source: University of California
Fast and efficient large content distribution is a challenge in the Internet due to its high traffic volume. In this paper, the authors propose ChunkCast, an anycast service that optimizes large content distribution. The authors present a distributed localityaware directory that supports an efficient query for large content. The system improves the median downloading time by at least 32% compared to previous approaches and emulates multicast trees without any explicit coordination of peers.