Epidemic and gossip algorithms attempt to spread information quickly through a network of processors. Traditionally, these algorithms have assumed that most processors and communications links run at approximately the same speed. This assumption is unrealistic for large Wide Area Networks such as the Internet that contain both slow and fast links and both slow and fast machines. The project will design and analyze highly-resilient and scalable mechanisms for propagating information rapidly that extend the epidemic approach to models that more accurately reflect real-world conditions.
