June 13, 2007
The New Network
A couple of weeks ago, I was in Washington DC. I met with a super-bright researcher at DARPA, and we hashed over some possible research ideas in secure service and data discovery. He suggested that my cohort and I read some of the latest offerings from Van Jacobson.
I've initially been a pretty stark critic of the idea of content-centric networking. My problem is, I guess, that I've read a lot of these SBIR proposals suggesting exactly this solution. The scale of CCN is hardly that of a six-month to one-year problem. Especially not in the context of cross-domain solutions, which generally need to have at least some semblance of assurance. Trying to use CCN as a content control mechanism is kind of a paradox. If the content is secured via a PKI, it's rather difficult for people to find it.
Jae used to have this cynicism about my cynicism concerning computer security. I always find it annoying to see new ideas put forth in the context of computer security, but which make some underlying assumptions about computer security which are patently false. So I think I'll shuck such thoughts to the wind for a while. Van Jacobson's Google Tech Talk is actually very cool -- he's a wizard at selling his idea, and I think it's a darned good one at that. As for a content control mechanism, I think it falls short in a lot of ways. It's kind of like ad-hoc networking with an assumption of pre-distributed shared keys (not very ad-hoc in my opinion). Still, I think the usefulness greatly outshines the limitations. As a content search and distribution system, it is The Way. Hopefully there are still some open areas of research here when I'm ready for my PhD...







by reid
on October 01, 2007
by reid
on July 17, 2005