The demand for IPv4 addresses has grown over the years since the Internet has evolved. As the demand increased, the pool of available addresses has been depleted and there is a shortage of available IPv4 addresses. Two major attempts to deal with this situation have become popular over time - NAT and IPv6.

NAT has become a fairly popular way to take a small subset of public IP addresses and utilize them across a larger number of devices on a network. NAT took off in the 1990s as a way to mitigate the IP starvation problem until IPv6 took off, but now it's so popular that many people don't want to let go of it. Meanwhile, system administrators are forced to embrace the looming legacy of NAT and make sure that we can tolerate it across our software and equipment due to popularity.

A new standard for IP addresses called IPv6 can solve the IP starvation problem by adding so many public IP addresses that we can have trillions of IP addresses for every square inch of the surface area of the Earth. We could give a block of IPs, the size of the entire Internet as we know it today, to every creature on the planet and not even put a dent in the total available pool of IPv6 addresses. The IPv6 specification was published in 1998 and has been slowly gaining momentum. IPv6 still takes a back seat to the more widely used IPv4 that was adopted in the 1970s. Most likely, even if we fully adopt IPv6, NAT still won't go away for some time.

The key problem we are trying to solve with NAT and VoIP is that since a device (phone) behind NAT is not visible to the Internet, it becomes difficult to contact that device when you want to call it. The next big problem is that some protocols, such as SIP, may break when used over NAT.