The Internet's Evolution: The Critical Shift from IPv4 to IPv6 and The Call to Action for Service Providers
Explore the transformative evolution of the internet in this latest feature. From its nascent stage in 2000 with 414 million users to a global network connecting over 5.35 billion individuals today, the digital landscape has undergone a seismic shift.
Articles published March 8, 2024
On Tuesday, March 26th, 2024, Jason Andersen will be featured as an expert speaker at the 2024 Iowa Communications Alliance Annual Meeting & Expo. His informative session will delve into the current state of IPv6 adoption and its significance for service providers, as well as the importance of protecting the client experience during the IPv6 adoption journey.
Introduction
In the year 2000, when broadband internet was still in its infancy, the World Wide Web had an estimated 414 million users worldwide – representing about 6% of the world’s population at that time.[1] Since that time, the internet has exploded. Today, the internet has an estimated 5.35 billion users – around 66% of the global population.[2]
This remarkable expansion owes much of its success to the steady stream of advancements made in fiber-optic broadband and cellular technology over the last two decades. But while service provider networks continue to evolve at a rapid pace, an often-neglected fact is that Internet Protocol Version 4 (IPv4) – the routing protocol upon which the internet was built and continues to operate – dates to the early 1980’s.
Despite IPv4’s impressive service record, exponential internet growth now forces both the telecommunications industry and governmental bodies across the world to confront its limitations. If ignored, the constraints of IPv4 now pose a significant threat to the ongoing integrity of the World Wide Web and the industry's ability to satisfy escalating consumer demand.
A Little History and The Origins of IPv4
Back in 1969, ARPANET – the world’s first wide-area, packet-switched network – was launched. It was established as a research project by a wing of the United States Department of Defense, and represented a ground-breaking stage of development toward what we now call the Internet.
When ARPANET began its life, it utilized a simple network control protocol, which within a few years was refined info Transmission Control Protocol (TCP) and Internet Protocol (IP). In 1980, the TCP standard was codified as the “DOD Standard,”[3] and IPv4 was published shortly thereafter.[4] Finally, in 1983, ARPANET and its interrelated networks made the official switch to TCP/IPv4, and the internet as we know it was born.
The Problem: Why Do We Need IPv6?
In order to communicate on a TCP/IPv4 network like the internet, a device requires a unique IPv4 address. Much like a home address, an IP address lets the rest of the world know how to reach your device. In simple terms, when you navigate to a website, your device sends packets to the destination website’s IP address, asking it to return information to your device’s IP address.
The problem has to do with scalability. IPv4 utilizes a 32-bit address field. This arrangement allows IPv4 to service around 4.3 billion unique addresses. Even before broadband internet came on the scene in the late 90’s, internet researchers had already foreseen that IPv4 could not sustain future growth.
As early as 1993, researchers were already coming up with ways to circumvent IPv4’s address limitation and extend its lifespan. In 1994, RFC 1631 codified Network Address Translation (NAT) as a method to allow IP addresses to operate in private networks “behind” a publicly routable IP address.[5] In 1996, the RFC 1918 standard was published, which set aside a specific range of IP addresses for use in private networks only.[6] Combined, these two developments dramatically mitigated the imminent threat of IPv4 address exhaustion, and have greatly shaped the way the internet is used today.
Without these developments, internet growth could have come to a screeching halt as early as the late 90’s. But even these developments were only stop gap solutions which did not address the larger issue of IPv4’s inherent limitations.
The IPv6 Solution Defined – But Not Deployed
In 1995, the long-term solution was introduced: IP version 6.[7] Improving upon IPv4’s 32-bit address limitation, IPv6 would leverage 128-bit addresses, resulting in a functionally limitless number of unique addresses.
However, by the mid-90’s, the world was already moving full steam ahead on IPv4. The World Wide Web was now alive, with tens of thousands of websites – all supported by IPv4. Operating systems, applications, and end user hardware were designed around IPv4. Internet service providers were already investing in IPv4-based infrastructure to provide service to subscribers who needed access to IPv4 content.
With the IPv4 train fully in motion, the world missed its window to make an easy pivot to IPv6. As a result, the internet would continue to grow on IPv4 for the next 20 years, and the issue of IPv4 exhaustion would be deferred as a future concern.
The Bill Comes Due
Despite efforts made to conserve IPv4 space, the Internet Assigned Numbers Authority (IANA) announced in 2011 that the world had officially run out of IPv4 addresses.[8] Later that year, the Asia-Pacific regional registry organization, APNIC, announced that their pools were exhausted.
With exhaustion now becoming a reality, an IPv4 resale market emerged that same year when in a first-of-its-kind transaction, Nortel Networks sold their IPv4 addresses to Microsoft for $7.5 million ($11.25 per address).[9] This served as a wakeup call to many service providers that the days of getting free IP space from your regional registry were soon ending.
In 2014, LACNIC (South America) announced their space was depleted, followed swiftly by ARIN (North America) in 2015. In 2019, RIPE (Europe) announced their final depletion.
Since the mid-2010’s, it has been clear to the world telecom community that the issue of IPv6 adoption is now an imminent concern. IPv4 prices have continued to rise since the market's inception in 2011, peaking in Q1 2022 at nearly $55 per address.[10] Pricing increase trends between 2011 and 2020 outpaced the Consumer Price Index, suggesting that value was driven by more than simple inflation. IPv4 address space is now a valued commodity – a precious resource required to connect subscribers to the internet.
IPv6 On the Rise
Since 2012, deliberate measures toward IPv6 adoption have been taken by many internet service providers and some world governments. However, change can only be accomplished gradually. Since most of the internet still runs on IPv4, providers moving forward with IPv6 deployment in their networks must take special measures to ensure their subscribers are able to reach both IPv4 and IPv6 content without negatively impacting client experience.
Google estimates that only 49% of internet subscribers in the United States are operating on IPv6-capable networks.[11] Cloudflare observes that about 36% of their total HTTP/HTTPS traffic is IPv6.[12] Similarly, APNIC estimates 36.5% of worldwide hosts to be IPv6-capable.[13] We have a long way to go, but steady progress has been made each year since 2012.
The biggest strides in IPv6 adoption have been made by wireless carriers, who enjoy certain inherent advantages. While the internet at large is comprised of countless end user devices of varying ages and capabilities, virtually all mobile devices designed in recent years are purpose-built with mechanisms to support easy transitioning to IPv6-only networks. As of 2021, T-Mobile was reported to have 91% of their subscriber base on IPv6, placing them on par with the Indian mobile giant Reliance Jio.[14]
In 2021, the Chinese government unveiled a very aggressive plan for IPv6 adoption.[15] By the end of 2025, all new networks in China are to be deployed as IPv6-only, and by the end of 2030, China intends for all networks to be fully converted to IPv6-only. If these ambitions are realized, it could tip worldwide IPv6 adoption measurements in an interesting way.
A Call to Action for Service Providers
Many US broadband providers are turning to Carrier-Grade NAT (CGNAT) solutions, which allow them to further stretch their IPv4 address pools by assigning multiple subscribers to the same IPv4 address, while simultaneously deploying IPv6. However, many US providers continue to operate on the status quo of issuing unique public IPv4 addresses to each subscriber. A10 Networks recently published a report which found that 39% of rural broadband providers still have no plans in place to transition to IPv6.[16]
With IPv4 exhaustion now a reality, service providers who are behind on embracing IPv6 are opening themselves to future business risk. As the world at large continues to adopt IPv6, the asset value of IPv4 will gradually decrease, which means non-IPv6 adopters who continued investing in IPv4 will eventually be left with nonvaluable assets. More importantly, the days are coming – perhaps sooner than we think – when the world will begin to see IPv6-only web content. If internet service providers aren’t ready for that, they run the risk of leaving their subscribers with portions of the internet inaccessible.
While it’s unlikely that IPv4 will be fully eliminated for many years, it's time for service providers to get serious about IPv6. The future of the internet depends on it.
Jason Andersen is a Product Manager with Aureon. He has been in the telecom industry for 17 years and has spent most of that time in various network and sales engineering roles. Jason has a broad range of networking experience having worked with ISP access networks, business VoIP and SDWAN solutions, as well as enterprise Ethernet and DWDM transport.
[1] Our World in Data, “Internet,” Accessed February 13, 2024, https://ourworldindata.org/internet
[2] Statista, “Number of internet and social media users worldwide as of January 2024,” January 31, 2024, https://www.statista.com/statistics/617136/digital-population-worldwide/
[3] Internet Engineering Task Force, “RFC 760,” January 1980, https://datatracker.ietf.org/doc/html/rfc760
[4] Internet Engineering Task Force, “RFC 791,” September 1981, https://datatracker.ietf.org/doc/html/rfc791
[5] Ibid, “RFC 1631,” May 1994, https://datatracker.ietf.org/doc/html/rfc1631
[6] Ibid, “RFC 1918,” February 1996, https://datatracker.ietf.org/doc/html/rfc1918
[7] Ibid, “RFC 1883,” December 1995, https://datatracker.ietf.org/doc/html/rfc1883
[8] Number Resource Organization (NRO), “Free Pool of IPv4 Address Space Depleted,” February 3, 2011, https://www.nro.net/ipv4-free-pool-depleted
[9] The Register, “Microsoft spends $7.5m on IP addresses,” March 24, 2011, https://www.theregister.com/2011/03/24/microsoft_ip_spend/
[10] IPv4.Global, “IPv4 Address Prices & Pricing,” Accessed February 13, 2024, https://ipv4.global/blog/ipv4-address-prices/
[11] Google, “Google IPv6,” Accessed February 13, 2024, https://www.google.com/intl/en/ipv6/statistics.html
[12] Cloudflare, “Cloudflare Radar,” Accessed February 13, 2024, https://radar.cloudflare.com/adoption-and-usage
[13] APNIC Labs, “IPv6 Capability Metrics,” Accessed February 13, 2024, https://stats.labs.apnic.net/ipv6/XA
[14] IPXO, “Detailed IPv6 Adoption Review,” September 10, 2021, https://www.ipxo.com/blog/detailed-ipv6-adoption-review/
[15] The Register, “China sets goal of running single-stack IPv6 network by 2023, orders blaze upgrade,” July 26, 2021, https://www.theregister.com/2021/07/26/china_single_stack_ipv6_notice/
[16] A10 Networks, “Insights 2024: Rural Broadband Business Sustainability,”