Routing in the Internet

[golde at cs.washington.edu (Hellmut Golde)]

Having just read Tom Narten's paper on Internet Routing, having
just started to look at Charles Hedrick's IGRP-ODV draft and the
OSPF spesification in RFC 1131, and having just briefly
looked at Hans-Werner Braun's paper on Border Gateway Protocol, I
cannot help wondering where the Internet (and all internets) are
going with respect to routing.

It appears to me that we need to take a comprehensive look at viable
topologies and routing stategies for large internets, perhaps
starting with the concept of an n-level (3-level a la NSFnet)
hierarchy.  

I assume that people who have been intimately involved in
running the Internet and its predecessors and offspring must have
worried about this problem, so I am sending this request to those
of you who know:

What is the current thinking about a "completely new" routing
structure for the Internet and for future, faster internets?
Who all is working on this problem?
Who is coordinating the effort?  Or is that a non-issue?

Any pointers are welcome.


Hellmut Golde
golde at cs.washington.edu