anyone know of research on live video over IP (Score 1)

Does anyone know of current research on increasing performance live video over IP? I don't know whether the designers of IP factored in high-density media, but with the universal adoption of TCP/IP it seems that we're "stuck"--not saying that IP isn't the best solution. Obviously we have a lot of practical applications of transferring recorded video--all the file sharing services, both web-based and client-based; BitTorrent. Are there any interesting technologies that currently exist or are on the horizon for streaming live video, that involves something really innovative, not just tweaking packet sizes and local app-layer state machines? I'm especially curious about mass media broadcasting--this goes against the net's "I only want to see what I ask for" pull behavior, but sometimes consumers are lazy and they actually want to be pushed. Maybe for pre-recorded TV programs live broadcasting isn't as important, but programs that are inherently live such as sports or news it will still be important. How is big media going to deliver relavent, live content? Will we still continue with cable, or do people see live broadcasts transitioning toward IP? If it is going the IP route, how? I know that multicast can be used on LANs...does anyone see media companies sectioning off networks, adding the infrastructure, and doing multicast on regional WANs?