Scalability is the main and toughest foe for system to grow large. What's the solution? --Distribution! P2P file sharing, web-caching... all these "hot" issues can be viewed as content distribution. The concept of "Content" may also be extened to any resources (including services). This means lots of things: content identification, content synchronization (replacement), content allocation, server placement... My current interests are placement, allocation and balance of the distributed content.
P2P-like Content Distribution Simulator
Peers Many concepts in p2p are from distributed database.
Pastry from Rice University
Chord of MIT -- Some friends in the area suggested me to see more in the easten coast.:)
Stanford Identiscape Their main purpose is not on content distribution, they are working on identification of mobile people. Anyway, some good ideas on p2p-like network in this group
Joltage A business solution on Ad-hoc wireless network
Ingrid This project focuses on information discovery
Grid at MIT PDOS Group
Iceberg Project at UC Berkeley I think this project focuses on the free communication between different kinds of devices more. Anyway, some of the basic mechanisms are very similar.
UND's self organized network (Biology View) a student's summary on some paper a student's summary on some paper