The current Internet design - Internet protocol version 4 (IPv4) - was standardized in 1978, and has 4.3 billion unique terminations. Simply put, "We can foresee a time when the allocations of addresses will end," said Cerf, who is known for his work in helping to create what we now know as the Internet.

The number of unique Internet addresses in IPv4, he said, will run out around 2011.

Enter IPv6, which has "about 340 trillion, trillion, trillion addresses - or 3.4 x 10³⁸," Cerf said. This exponential number of IP addresses means more mobile devices, which helps in emergency management.

Q: How does IPv6 create more mobility?
A: IPv6 and IPv4 are both relatively easily used to dynamically create new networks. There's something called DHCP [Dynamic Host Configuration Protocol], which is a way in which you connect to the Net, and you're assigned an IP address. So building emergency networks quickly is readily done in both IPv4 and IPv6 contexts.

Mobility - movement from one IP address to another - if you disconnect from the Net, maybe you radio disconnect or physically disconnect, and you plug in somewhere else, you often get a different IP address.

The current protocols to the Internet are not as friendly to that kind of mobility as they could be, so there's some serious technical work that could be done, and I think should be done by the research agencies and universities to make the protocols and the network more comfortable with mobility. [Disruption Tolerant Networks] DTN is an example of that - where it's assumed that things will be disrupted as opposed to being mostly connected and occasionally disrupted.

So dynamic networking and the so-called mobile access network that self-organize, are very important for emergency services, networking and sensing types of services. I think we can achieve our objectives with IPv4 or IPv6 - IPv6 just has the benefit that there is more address space available.

Q: In what other ways is IPv6 useful to the emergency management field?
A: I think there's another element here that might not be quite as obvious. In many emergency communication systems, radio compatibility is required for the different emergency responders to communicate with each other. So in the absence of compatible radios, you find voice communication is extremely awkward to impossible.

If we were to move to a voice over IP architecture then the end-to-end communication would be digitally formatted and use basic Internet protocols. We still have to deal with the gap between this frequency radio and that frequency radio, but if we build systems that will communicate at both frequencies and simply relay the packets back and forth, that's not too different than what we used to call a gateway between two networks that's pulling something out of the low-level network format, extracting an Internet packet from it, embedding it in the next network's format and forwarding it on.

The hosts, or the computers at both ends, are communicating end-to-end through IP; they're not conscious of the fact that there were several different formats in which these packets were embedded, or they were conscious of the fact that there were several different frequencies that you hopped between.

So if we move to a voice over IP architecture, I think we have a huge opportunity to make the compatibility among all of the emergency services parties much easier because they'll be able to communicate with voice over IP. Of course, if they're running Internet protocols, they also have the ability to do a lot of other things together.

Q: Can you give some specific examples?
A: They can