Is there any ways to improve the IPv6 addressing types and size so that could improve the performance of internet to meet the future demand of internet users?
Yes, no, maybe.
Yes, an improvement is always possible.
No, IPv6 are so many and so flexible that there's no real need for any improvement.
Maybe. Maybe your question is missing something.
The main issue, for real, is the question. What do you mean with "meet the future demand of internet users" ? Define what is the "user" and what are the needed improvement (define a scenario where the improvement is needed). Then the solution might be found.
Performance-wise, IPv6 is as fast as IPv4, and there's a lot of space for fine-tuning. IPv6 "longer" header isn't really an issue, etc. Addressing scheme usually involves a *minimum* /64 per-user, and the types are extremely flexible.
Hence, in order to discuss about the topic (which might be as well be very interesting), we need to find a scenario where the current addressing scheme and type is not fitting well.
User means,I mean the increase in world's population who requests internet access for the next 50 Years.
let's do some simple math. Let's assume to give a "basic" /64 allocation scheme to every customer premise. Current global addresses are extracted from the 2000::/3 pool (see http://www.iana.org/assignments/ipv6-address-space/ipv6-address-space.xml). This leaves 61 bits for the NetId. More or less 2.3*10^18.
The world estimate population in 2100 will be 160 billion (highest estimate, according to https://en.wikipedia.org/wiki/World_population), i.e., 1.6*10^11.
This leads to (roughly) 10^7 networks for each human being, including all his/her devices, and so on.
Moreover, there are some "small" blocks currently unassigned by IANA. E.g., 4000::/3, 6000::/3. etc.
Hence, if everybody, no matter the age (even the infants) would like to surf the net though IPv6, at least until 2100 we're not going to run out of addresses.
Of course this is a really simple analysis, as institutions usually ask for a /48. However the unassigned blocks are there, and they're quite large.
Anyway, issues might be around the corner (as usual), so any research in this field is always welcome. I'd suggest to read this as a starter: http://www.ripe.net/ripe/docs/ripe-589
- Badges
- Science topic