History of This Work and the Term “Third Internet”
This book is an update and expansion of my 2010 ebook, The Second Internet. That ebook has been available on the main website of the global IPv6 Forum ( http://ipv6forum.com ) since 2010 with some 500,000 downloads worldwide. This book is actually still about the new Internet based on IPv6, but since 2010 I have realized that the ARPANET 1 is not phase 1 of the First Internet; it IS the First Internet. That makes the Internet based on IPv4 (still what most people are using today) the real Second Internet, which makes the new Internet being created now, based on IPv6, the Third Internet.
One notable change since 2010 is that IPv6 is no longer just a Draft Proposed Standard. The Official IETF Standard for IPv6 has finally been released 2 – RFC 8200 3: “Internet Protocol, Version 6 (IPv6 ) Specification,” July 2017 (STD 86). This replaces RFC 2460 and several additions to it. So you need to get used to referring to RFC 8200 instead of RFC 2460!
The leap from the First Internet (ARPANET ) to the Second Internet (IPv4 based) was clearly a generational change:
The foundation protocol of the First Internet was usually referred to as NCP,4 but officially was called the “host-host”5 protocol. It was defined in a few RFCs. Many new RFCs (starting with RFC 791 6 in 1981) specified the new IPv4 and related protocols.
An IPv4-only node could not make a connection to, or exchange information with, an NCP-only node (and vice versa), without a complex gateway. The inability to interoperate often happens in generational changes. Translation between NCP and IPv4 was never accomplished.
NCP had 8-bit addresses (max 28 or 256 addresses), while IPv4 has 32-bit addresses (max 232 or 4.3 billion addresses). That is four times as many bits in each address as in NCP, but 224 (16.7 million) times as many addresses. Each additional bit doubles the number of addresses.
The First Internet lasted from 1969 until 1982 with significant growth and evolution during those years. The Second Internet began operation on January 1, 1983, grew extremely rapidly, and is still running, although it is developing more and more serious issues related to exhaustion of the IPv4 public address space.7
While applications such as email, remote terminal emulation, and file transfer existed in the First Internet, all such apps had to be rewritten (significantly) to work over IPv4.
Engineers familiar only with NCP had to go back to the books (and training classes) to master the new IPv4. All software and hardware devices that worked with NCP had to be rewritten to work with IPv4. There was no “dual-stack” period since that transition was done via a “flag day” (only NCP before January 1, 1983, only IPv4 from then on). There were many serious problems with doing such an abrupt transition, like worldwide email broke for several months. The IETF wisely decided to do a more gradual transition from IPv4 to IPv6.
NCP node addresses were represented as a single one- to three-digit decimal number (e.g., “10”), while IPv4 addresses were represented using dotted decimal (e.g., “123.45.67.89”), which at the time looked very alien to NCP users.
A block diagram maps the entire first internet circa 1982. A R P A N E T 10 is the source from where the map is scattered among different networks.
Map of the entire First Internet circa 1982
An illustration depicts the map of the second internet that was proposed in 2015. The signals can be differentiated by paths against a dark background.
One proposed map of the Second Internet , 2015
The foundation protocol of the Second Internet (IPv4) was defined in several RFCs. Many new RFCs (starting with RFC 1881 in 1995) specified the new IPv6 and related protocols.
An IPv6-only node cannot make a connection to, or exchange information with, an IPv4-only node (and vice versa), without a complex gateway. One solution is “dual stack,” where every node has both IPv4 and IPv6 and hence can make connections to both the Second and Third Internets. Another is to run only IPv6 internally and provide access to external legacy (IPv4-only) nodes via a NAT64 8 gateway.
IPv4 has 32-bit addresses (max 4.3 billion values), while IPv6 has 128-bit addresses (max 340 trillion, trillion, trillion values). This is again four times as many bits as in IPv4, but now 296 times as many addresses as in IPv4. If you think of the IPv4 address space as the size of a basketball, the IPv6 address space is a sphere that would not only include the entire sun but go most of the way out to Venus. That’s a big ball.
The Second Internet began operation in 1983 and will probably be mostly phased out 9 by 2028 or so. There are no more public addresses for the Second Internet to grow with – all growth of the Second Internet today is in private Internets (networks that use RFC 1918 private addresses and are not directly connected to the public IPv4 Internet). Each of these private Internets is hidden behind an existing public IPv4 address with NAT 10 (or even behind multiple layers of NAT 11).
While applications such as email, remote terminal emulation, and file transfer existed in the Second Internet, most applications (aside from web scripts) must be rewritten to at least some extent to work over IPv6 (or, more commonly, over both IPv4 and IPv6). Since IPv6 has no NAT but ample global addresses, there are entirely new types of connections possible, such as servers on phones or end-to-end direct (e.g., connecting directly from my phone to yours, with no intermediary server).
Engineers familiar only with IPv4 are having to go back to the books and training classes to master the new IPv6. Engineers and developers who don’t learn IPv6 will find it more and more difficult to remain employed, like NetWare engineers experienced once the transition to TCP/IP (Transmission Control Protocol/ Internet Protocol)–based networks took place. If you know IPv4 today, this book contains enough technical detail on IPv6 to get you well along your way to mastering IPv6. I have helped many senior network and telco engineers make the leap to IPv6 as a gold-certified IPv6 Forum trainer.
IPv4 addresses were represented using dotted decimal notation (e.g., “123.45.67.89”), while IPv6 addresses are represented with what I call coloned hex notation (e.g., “2001:db8:ed3a:1000::2:1”), which looks very strange indeed to IPv4 engineers.
Why IPv6 Is Important
The Second Internet (aka the Legacy Internet) is now 36 years old. Think about what kind of CPU, amount of RAM, and which operating system you were using in 1983 – probably a Z80 8-bit CPU with 64 kilobytes of RAM and CPM/80 or, if you were a businessman, an 8088 “16-bit” CPU and MSDOS 1.0. If you were really lucky, you might have had an expensive hard disk drive with a massive 10 megabytes of storage. What, many of you reading this weren’t even alive then? Ask your father what personal computing was like in 1983. I’ve been building, programming, and applying personal computers since my Altair 8800 in 1975. Hard to realize that is 44 years ago. Since 1983, network speeds have increased from 10 Mbps to 100 Gbps (10,000-fold increase). Access from home may have been 1200 baud (1.2 kbps) then, but 100 Mbps to 1 Gbps today. Amazingly we are still using essentially the same Internet Protocol. Think it’s about time for an upgrade?
The Second Internet has impacted the lives of billions of people. It has led to unprecedented advances in computing, communications, collaboration, research, and entertainment (not to mention time-wasting, dating, gossiping, and even less savory activities). The Internet is now understood to be highly strategic in every modern country’s economy. There are now people claiming that access to the Internet is a “human right.” 12 It is difficult to conceive of a country that could exist without it. Many enormous companies (such as Google) would not have been possible (or even needed) without it. Staggering amounts of wealth have been created (and consumed) by it. It made “snail mail ” (paper mail physically delivered) follow the Pony Express into oblivion (amazingly, governments everywhere are still trying to keep post offices going, even though most of them lose gigantic amounts of money every year and they mostly only deliver advertising circulars). The number of emails sent daily is at least four times the number of first-class mails sent annually (in the United States).
Estimates are that there are about 26 billion nodes 13 (computers, servers, or other network devices) connected to the Internet as of 2019. Neat trick for a protocol with only 4.3 billion theoretically possible unique addresses, eh?
But wait. There’s more.
If you think that’s impressive, wait until you see what its rapidly approaching successor, the Third Internet (made possible by IPv6), will be. One estimate (same link as earlier) predicts some 75 billion nodes by 2025. Entirely new and far more flexible communication and connectivity paradigms are coming that will make email and texting seem quaint (e.g., 5G 14). Major areas of the economy – such as telephony, entertainment, and almost all consumer electronic devices (MP3 players, TVs, radios) – will be heavily impacted or even collapse into the Third Internet as just more network applications (like email and web did in the First and Second Internets). The Second Internet (the one you are likely using today, based on IPv4) that you think is so pervasive and so cool is tiny compared to the potential size of the Third Internet. One of the popular terms being used to describe it is pervasive computing .15 That means it is going to be everywhere, even inside your body (embedded sensors communicating via a Personal Area Network (PAN) using your phone as a relay to the Third Internet).
Flash! The Second Internet is broken!
Most importantly, in the process of keeping IPv4 around way too long, we’ve already broken the Second Internet badly with something called NAT 16 (Network Address Translation – much more on this later). NAT has turned the Internet into a one-way channel, introduced many major security issues, and is impeding progress on newer applications like Voice over Internet Protocol (VoIP) and Internet Protocol Television (IPTV) .
NAT has fragmented the old monolithic pre-1995 Second Internet into millions of private Internets, each hiding behind one public IPv4 address. You can easily make outgoing connections from your node to servers like www.facebook.com , but it is difficult or impossible for other people to make connections to your node. NAT has divided the world into a few producers (like www.facebook.com ) and millions of consumers (like you). You can post some content to their sites, but they own the sites and have complete control over what you can post and can withdraw your right to post at any time, for any reason.
In the Third Internet, anyone can be a prosumer (producer and consumer) of content. You will be able to run any server or even a global TV network from your node. You will be able to connect directly from your node to anyone else’s node in the world (assuming no firewalls block that connection). There is no shortage of public addresses – we can all be first-class netizens. NAT was a necessary evil to keep the Internet on life support until the Third Internet was ready to be rolled out. The transition to IPv6 was supposed to be finished by 2010. NAT has now served its purpose and, like crutches when your broken leg has healed, should be cast aside. Its only purpose was to extend the life of the IPv4 address space while the engineers were getting IPv6 ready. IPv6 IS ready and in rapid adoption mode globally now. This book should be a “wake-up call” for everyone using the Internet.
Using a “horses and cars” metaphor, there is no reason to wait for the last horse to die (the last IPv4 node to be shut down) before we start driving cars (deploy IPv6). Another aspect of that is we no longer need horse doctors; we need car mechanics! Good news, everyone! IPv6 is ready for prime time today. My home is already fully migrated to dual stack (IPv4 + IPv6). It has been for over a decade.
Wait. How Can the Internet Grow to 75 Billion Nodes?
If there are only about 7.5 billion people alive, how can the Internet possibly grow to 75 billion nodes ? The key here is to understand that the Third Internet (based on IPv6) is the Internet of Things.17 A human sitting at a keyboard will be a relatively rare thing. However, IPv6 will make it far easier and cheaper to bring the next billion humans online using IPv6’s advanced features and almost unlimited address space. Many Asian countries and companies (who routinely have 5- to 10-year horizons in their planning) already consider IPv6 to be one of the most strategic and important technologies anywhere and are investing heavily in deploying IPv6. 2018 was the tipping point 18 for IPv6. Adoption curves are starting to climb at steep rates reminiscent of the adoption of the World Wide Web back in the mid-1990s. By 2018, more than 50% of all global traffic was over IPv6 in many countries. IPv4 will be in decline, with worsening service and fewer and fewer public addresses, at any price. Also, today, many people have multiple devices connected to the Internet. In Singapore, 23% of the people have five or more nodes with Internet connectivity.
Why Was 2011 a Significant Year for the Second Internet?
There is an entire chapter in this book on the depletion of the IPv4 address space. What this means (in English) is that we are running out of public IPv4 addresses for the Second Internet. On February 3, 2011, there was a very important event in the history of the Internet. I woke my kids up to watch it live-streamed over the Internet, so they could tell their kids that they saw the beginning of the end of the IPv4 Internet. IANA allocated the final five unallocated blocks of IPv4 public addresses to the five RIRs (Regional Internet Registries ).
In the mid-1990s, the folks in charge of the Internet realized we would soon run out of public IPv4 addresses and only managed to keep the Internet going through some clever tricks (NAT and private addresses), kind of like using private extension numbers in a company PBX phone system. However, even with this trick (which is now causing major problems), we have pretty much run out for good. All the groups that oversee the Internet – like the Internet Assigned Numbers Authority (IANA 19), the Internet Corporation for Assigned Names and Numbers (ICANN 20), the Internet Society (ISOC 21), the Internet Engineering Task Force (IETF 22), and the Regional Internet Registries (RIRs 23) – have been saying for some time that the world has to migrate to IPv6 now. The five Regional Internet Registries are ARIN,24 RIPE NCC,25 APNIC,26 LACNIC,27 and AfriNIC.28 They should know. They are the ones that manage and allocate public IP addresses to telcos, ISPs (Internet Service Providers), and cloud providers. They know that the IPv4 barrel is pretty much empty. We’ve got to provide tens of billions more globally unique Internet addresses, which has some far-reaching consequences. There is no additional source of IPv4 addresses, so these will have to be IPv6 addresses.
An Analogy: The Amazing Growing Telephone Number
When I was very young, my family’s telephone had a five-digit phone number (let’s say it was 5-4573). That covered only my small town (about 10,000 people at the time). As the number of phones (and hence unique phone numbers within my geographic region) grew, the telephone company had to increase the length of everyone’s phone number. Our number became 385-4573 (seven digits), enough for 107 (10 million) phone numbers. This was enough to give everyone in my part of Florida a unique number, and we could ask the nice long-distance operator to connect us to people in other areas when we wanted to talk with them. When the telcos introduced the miracle of Direct Distance Dialing, our phone number grew to ten digits by adding an area code: for example, (904) 385-4573. In theory, this could provide unique numbers to 1010 (10 billion) customers. In practice some digit patterns cannot be used, so it is somewhat less than that, and today many people have multiple phone numbers (landline, cell phone, fax, modem, VoIP, etc.). Estimates are that the current supply of ten-digit numbers will last US subscribers at least 50 more years. Increases in the length of phone numbers may be an inconvenience to end users (and publishers of phone books), but the tricky problems are mostly in the big telephone company switches. Phone number lengths have been increased several times over the years, without leading to the collapse of civilization.
With 5G, numeric (aka E.164 29) phone numbers are going away. In the future, your “phone number” will look like sip:[email protected]. There are an essentially unlimited number of SIP URIs (Uniform Resource Identifiers).30 They are also conveniently organized into the same hierarchy used for email and web.
At the top (IANA) level, the final five unallocated blocks of IPv4 public addresses (16.7 million each) were given out to the five Regional Internet Registries 31 on February 3, 2011.32 Since that date, if the RIRs asked for additional blocks of IPv4 addresses, IANA would tell them, “Sorry. The cupboard is bare.” The RIRs had enough on hand to last a while, but those are gone now (except for Africa). I once bought some addresses from APNIC as a member and reserved a “/22” block of IPv4 addresses (a little over 1000 of the precious, and increasingly scarce, addresses for the Second Internet). These cost me about 1000 USD per year, but I could have used those for many things. You can think of this as staking out some of the last remaining lots in a virtual Oklahoma Land Rush. At the same time, I got my very own “/32” block of shiny new IPv6 addresses. You can think of this as getting an enormous spread of prime real estate in the virtual New World of the Third Internet. A few years ago, I got tired of paying the charges and returned those blocks to APNIC. I now have one public IPv4 at home, which I had to pay $50 for when I signed up with my current ISP and can keep so long as I have service with them (very effective marketing – if I gave this one up, it is unlikely I would ever get another).
There is a flourishing “gray market” for IPv4 addresses today. Going rate is about $16 per public IPv4. That price will go up until IPv6 is widely deployed, at which point that price will drop to zero quickly.
So Just What Is It That We Are Running Out Of?
There is a great deal of confusion and misunderstanding about this issue, as important as it is. Many people think that an “Internet address” is something like www.ipv6.org . That is not an Internet address; that is a domain-qualified symbolic nodename. That is an important part of a URI (Uniform Resource Identifier ), which adds things such as a protocol designator (e.g., http:, mailto:, or sip:), possibly a nonstandard port number (e.g., “:8080”), and often a file path (e.g., “/files/index.html”). There are still a staggering number of possible domain-qualified nodenames that are easy to remember, more than could ever be used in the next hundred years. So just what is it that we are running out of?
The nodenames that you (and most humans) use to specify a particular node on the Internet, like www.ipv6.org , are made possible by something called the Domain Name System (DNS 33). Those nodenames are not used in the actual packets as source and destination addresses (see the section on the IPv4 addressing model for the gory details). The addresses used in the packets in the Second Internet are 32-bit binary numbers. These are usually represented for us slow and stupid humans in dotted decimal notation like 123.45.67.89. With a 32-bit address, there are 232 (about 4.3 billion) distinct values. When you use a symbolic nodename (known technically as a fully qualified domain name, or FQDN) in an application, that application sends it to a DNS server, which returns the numeric IP address associated with it. That’s the address that is used in packets on the wire, for routing the packet to its destination.
The DNS nodenames are like the names of people you call; the IP addresses are like their phone numbers. DNS is like an online telephone book that looks up the “phone number” (IP address) for “people” (nodes) you want to “call” (connect to). Did you know that you can surf to an IP address? Try entering the URL http://15.73.4.75. That’s a whole lot harder to remember than www.hp.com , which is why DNS was invented. It’s these 32-bit numeric addresses (that most people never see) that we are running out of. The good news is that you can keep typing www.hp.com and DNS will return both the old-style 32-bit IPv4 address and a new-style 128-bit IPv6 address, which will be put into the network packets. Given the choice, your applications will prefer to use the new IPv6 address. You will hardly notice the difference unless you are a network engineer or a network software developer, except there’s going to be a whole bunch of cool new stuff to do and new ways of doing old things. Plus, the Internet is going to work better than it ever has before.
Can you imagine trying to use telephones today with five-digit telephone numbers? In a few years, that’s what IPv4 is going to feel like. I’ve been using IPv6 for over a decade, and IPv4 already looks antiquated to me. It’s amazing we were able to build the current Second Internet with something so primitive and limited. I’m creating new apps for IPv6 already.
But You Said There Were 4.3 Billion IPv4 Addresses?
There are 26 billion nodes connected to the Second Internet, but only 4.3 billion IPv4 addresses ? How does THAT work? Well, there are probably around 3 billion usable IPv4 public addresses (and essentially no new ones to allocate, except in very specific circumstances, like for IPv6 migration). The bulk of those nodes are not on the public IPv4 Internet, but in private Internets hiding behind NAT gateways. Pretty much all new nodes being added (like all those cell phones that can connect to the Internet) are in private Internets. There is no real shortage of addresses for private Internets. In theory, every public address could have as many as 16 million nodes (the number of possible nodes in the 10/8 private subnet) behind it. In practice a single NAT gateway can’t handle anywhere near that many nodes, but there can still be hundreds or even thousands of nodes in each private Internet. The problem is that nodes in private Internets can’t accept incoming connections, except via NAT traversal 34 (which introduces many security issues). NAT also breaks a lot of important protocols, like VoIP and IPsec. NAT was only ever meant as a temporary stopgap measure during the transition to IPv6. A lot of people today (including telcos and ISPs) seem to think we can just go on using IPv4 with NAT forever. We can’t.
How did we get into this situation? Well, when the Second Internet was being launched, there were about 200 nodes on the First Internet, and 4.3 billion looked a lot like “infinity” to the people involved. So giant chunks of addresses were generously given out to early adopter organizations. For example, MIT and HP were given “class A” blocks of addresses (about 16.7 million addresses each, or 1/256 of the total address space). Smaller organizations were given “class B” blocks of addresses (each having about 65,535 addresses). Most of these organizations are not using anywhere near all those addresses, but they have only rarely been willing to turn them back in to be reallocated to newcomers. As detailed in the Organisation for Economic Co-operation and Development (OECD) study on IPv4 address space depletion and migration to IPv6, it is very difficult and time-consuming to recover these “lost” addresses. Also, some blocks of IPv4 addresses were used for things like multicast (“class D”), experimental use (“class E”), and other purposes like addresses for private Internets (RFC 1918 35).
We are getting more efficient in our allocation of blocks of IPv4 addresses, but even with every trick we know, they are all gone now at the top (IANA) level and four of the five RIRs (ARIN, APNIC, RIPE, and LACNIC). There are something like 1.5 billion smartphones being sold each year (and this doesn’t even count other devices that might need addresses). There may be tens of billions of IoT nodes. How do we connect all these? This can only be done by going to longer IP addresses (hence, a larger address space). This is one of the main things that IPv6 is about.
Is IPv6 Just an Asian Thing?
Some time ago, I heard some comments from US networking professionals and venture capitalists that IPv6 was an “Asian thing ,” something that is of little interest or concern to Americans. This shows an unusually provincial view of an extremely serious situation. This attitude was only partly due to the inequitable distribution of addresses for the Second Internet (there are over six IPv4 public addresses per American citizen, compared with only about 0.28 per person for the rest of the world). It has a lot more to do with a lack of knowledge of how certain parts of the Second Internet really work, compounded by a limited time horizon compared with Asian businessmen, who routinely plan 5–10 years ahead. American business schools teach that nothing is important beyond the next quarter’s numbers. The depletion of IPv4 addresses is already here. Some American businessmen are now panicking (“Why didn’t you warn us about this?”).
Since 2010, US mobile telephone service providers have embraced IPv6 enthusiastically, more so than other regions or industries. They realized they could deploy only IPv6 for a far lower cost than trying to keep IPv4 alive one more year. Also, it was becoming a challenge to knit together multiple /8 subnets (the largest you can create with IPv4 private addresses), each of which is 16.7M addresses. Many telcos have far more than 16.7M customers. With IPv6 there is no such problem. Now that all Android phones include 464XLAT,36 even legacy IPv4-only mobile apps work just fine. On iOS, Apple requires that apps work in an IPv6-only environment before they are approved for the App Store.
Any country or organization that (for whatever reason) doesn’t migrate to IPv6 is going to still be “riding horses” while the rest of us are zipping around in these newfangled “cars.” When I wrote the 2010 version of this book, I was having nightmares about the United States being just as reluctant to go to IPv6 as they were to adopt the metric system (the United States is the only industrialized country not to have adopted the metric system, and I doubt they ever will). They could have decided to stay with IPv4. If they had, it would have become increasingly difficult for them to connect to non-US websites or for people in other countries to connect to US websites. It would have impacted all telephone calls between the United States and anywhere else in the world. It would have made IT products designed for the US market of little interest outside of the United States (kind of like automobiles that can’t be maintained with metric tools). This would have isolated the United States even further and essentially leave leadership in Information Technology up for grabs. Japan, China, and South Korea are quite serious about grabbing that leadership, and they are well along their way to accomplishing this, by investing heavily in IPv6 since the late 1990s. Since then, America has finally “gotten religion” about IPv6, especially in mobile telephone service providers where IPv6 is approaching 100%.
Being good engineers, while the IETF has the “streets dug up” increasing the size of IP addresses, they fixed and enhanced many of the aspects of IPv4 (QoS, multicast, routing, etc.) that weren’t done quite as well as they might have been (who could have envisioned streaming video 34 years ago?). IPv6 is not just bigger addresses. It’s a whole new and remarkably robust platform on which to build the Third Internet.
So Exactly What Is This “Third Internet”?
Most things in computer technology evolve through various releases or generations, with significant new features and capabilities in the newer generations, for example, 2G, 3G, and 4G cell phones. The Internet is no exception. The remarkable thing, though, is that the Second Internet has lasted for 36 years already. The third generation has been quietly emerging for some time and is now well underway. 5G phones will be mostly based on IPv6. There are many technology trends going on right now, and some of them have been hyped heavily in the press. Some of them sound a lot like they might be the next generation of the Internet. Let’s see if we can narrow down what I mean by “the Third Internet” by discussing some of the things that it is not.
Is It the Next-Generation Network (NGN) That Telcos Talk About?
Telcos around the world have been moving toward something they call NGN 37 for some time. Is that the same thing as the Third Internet? Well, there is certainly a lot of overlap, but, no, NGN is something quite different.
Historically, telephone networks have been based on a variety of technologies, mostly circuit switched, with call setup handled by SS7 38 (Signaling System 7 ). The core of the networks might be digital, but almost the entire last mile (the part of the telco system reaching from the local telco office into your homes and businesses) is still analog today. There was some effort at upgrading this last mile to digital with ISDN 39 (Integrated Services Digital Networks ), but some terrible decisions regarding tariffs (the cost of services) pretty much killed ISDN in many countries, including the United States.
The ITU 40 (International Telecommunication Union ), an agency of the United Nations that has historically overseen telephone systems worldwide, defines NGN as packet-switched networks able to provide services, including telecommunications, over broadband, with Quality of Service (QoS) –enabled transport technologies, and in which service-related functions are independent from underlying transport-related technologies. It offers unrestricted access by users to different telecommunication service providers. It supports generalized mobility, which will allow consistent and ubiquitous service to users.
In telco core networks, there is a consolidation (or convergence) of legacy transport networks based on X.25 and Frame Relay into the data networks based on TCP/IP (some still using IPv4, but more and more core networks are IPv6 today). It also involves moving from circuit-switched (mostly analog) voice technology (the Public Switched Telephone Network, or PSTN 41) to Voice over Internet Protocol (VoIP 42). So far, the move to VoIP is mostly internal to the telcos. What is in your house and company is good old POTS 43 (Plain Old Telephone Service).
In the “last mile,” NGN involves migration from legacy split voice and data networks to Digital Subscriber Line (DSL) , making it possible to finally remove the legacy voice switching infrastructure. Today, more and more telcos are running FTTH 44 (Fiber to the Home), which is of course digital all the way.
In cable access networks , NGN involves migration of constant bit rate voice to Packet Cable standards that provide VoIP and Session Initiation Protocol (SIP) services. These are provided over DOCSIS 45 (Data Over Cable Service Interface Specification ) as the cable data layer standard. DOCSIS 3.0 does include good support for IPv6, though it requires major upgrades to existing infrastructure. There is also a “DOCSIS 2.0 + IPv6” standard, which supports IPv6 even over the older DOCSIS 2.0 framework, typically requiring only a firmware upgrade in equipment. That will likely get rolled out before DOCSIS 3.0 can be. Especially in the United States, DOCSIS 3.0 is finally being widely deployed, with speeds even above 1 Gbps.
A major part of NGN is IMS 46 (the IP Multimedia Subsystem ). To understand IMS, I highly recommend the book The 3G IP Multimedia Subsystem (IMS) : Merging the Internet and the Cellular Worlds, by Gonzalo Camarillo and Miguel A. Garcia-Martin. This was published by John Wiley & Sons, in 2004. This book says that IMS (which is the future of all telephony) was designed to work only over IPv6, using DHCPv6, DNS over IPv6, E.164 Number Mapping (ENUM), and Session Initiation Protocol/Real-Time Transport Protocol (SIP/RTP) over IPv6. IMS is so IPv6 specific that some of the primary concerns are how legacy IPv4-only SIP-based user agents (hardphones and softphones) will communicate with the IPv6 core. One approach is to use dual-stack SIP proxies that can in effect translate between SIP over IPv4 and SIP over IPv6. Translation of the media component (RTP) is a bit trickier and will be handled by Network Address Translation between IPv4 and IPv6. Newer IPv6-compliant user agents will be able to interoperate directly with the IMS core, without any gateways, and solve many problems. They are beginning to appear. One example is some dual-stack IP phones from the Korean company Moimstone.47
The first “Internet over telco wireless service” in early 2G networks was WAP 48 (Wireless Application Protocol ). WAP 1.0 was released in April 1998. WAP 1.1 followed in 1999, followed by WAP 1.2 in June 2000. The Short Message System (SMS 49) was introduced, but only IPv4 was supported. Speed and capabilities were somewhat underwhelming.
2.5G systems improved on WAP with GPRS 50 (General Packet Radio Service ), with theoretical data rates of 56–114 Kbps. GPRS included “always on” Internet access, Multimedia Messaging Service (MMS 51), and point-to-point service. It increased the speed of SMS to about 30 messages/second. Even Filipinos can’t text that fast. As with WAP, only IPv4 was supported.
2.75G systems introduced EDGE 52 (Enhanced Data Rates for GSM Evolution ), also known as EGPRS (Enhanced GPRS). EDGE service provided up to 2 Mbps to a stationary or walking user and 348 Kbps in a moving vehicle. IPv6 service has been demonstrated over EDGE but is not widely deployed.
3G systems introduced HSPA 53 (High-Speed Packet Access ), which consisted of two protocols, HSDPA (High-Speed Downlink Packet Access ) with theoretical speeds of up to 14 Mbps service and HSUPA (High-Speed Uplink Packet Access ) with up to 5.8 Mbps service. Real performance was again somewhat lower, but better than with EDGE. HSPA had good support for IPv6.
The last gasp for 3G (sometimes called “3.9G”) is LTE 54 (Long-Term Evolution ). LTE is completely based on IP and was supposed to be based on IPv6. Early versions of the specification clearly described it with IPv6 mandatory and IPv4 support optional. It was later reworded to make most aspects “IPv4v6” (dual stack). The reality is mostly just IPv4. 3G was still based on two parallel infrastructures (circuit switched and packet switched). LTE is packet switched only (“all IP”). There are a few deployments of LTE (some of which are described incorrectly as “4G”) around the world.
4G systems have been around for some time. These provide even higher-speed wireless transports. Originally 4G was supposed to be the big change to IP only, but IPv6 wasn’t widely enough deployed, and vendors wanted to sell the higher speed as something really different.
So 5G is now being deployed. This will use an all-IP infrastructure for both wired and wireless. The specification for 5G claims peak downlink rates of as much as 1 Gbps and uplink rates of several hundred Mbps. 5G requires a “flat” IP infrastructure (no NAT), which can only be accomplished with IPv6. IPv4 address space depletion happened some time ago, so IPv4 is not even an option this time around. IPTV 55 is a key part of 5G, which requires fully functional multicast, scalable to very large customer bases. That also requires IPv6.
So clearly the telco’s NGN is moving more and more toward IPv6. Some deployments are still mostly IPv4. However, NGN is just as clearly not the Third Internet described in this book. You might say that NGN (once it reaches 5G) will be just another one of the major subsystems hosted on the Third Internet, peer to email, the Web, IPTV, etc. 5G is also called “the Grand Convergence,” referring to the long-awaited merging of “the Internet” and “telephony” into a single seamless network.
The block diagram depicts the next-generation network. It includes the N G N services and N G N transport. The N G N services include video, data, and telephone services.
NGN
Is It Internet2 or National LambdaRail?
Internet2 59 is an advanced academic and industrial consortium led by the research and education community, including over 200 higher education institutions and the research departments of several large corporations. They have deployed a worldwide research network called the Internet2 network. While IPv6 is definitely being used on Internet2, they also use a lot of IPv4. Their focus is more on very high performance than which version of IP is used. The first part of the Internet2 network (called Abilene 60) was built in 1998, running at 10 Gbps, even over Wide Area Network (WAN) links. It was associated with the National LambdaRail 61 (NLR) project for some time. Internet2 and NLR have since split and moved forward along two different paths. Today, most links in the global Internet2 network are running at 100 Gbps. This is 10–100 times faster than typical WAN links used by major corporations today.
Internet2 also features advanced research into secure identity and access management tools, on-demand creation and scheduling of high-bandwidth, high-performance circuits, layer 2 Virtual Private Networks (VPNs), and dynamic circuit networks (DCNs).
A recent survey of Internet2 sites showed that only a small percentage of them have even basic IPv6 functionality deployed, such as IPv6 DNS, email, or VoIP over IPv6. IPv6 is independent of their goals. Essentially, Internet2 is primarily concerned more with extreme high-end performance (100 Gbps and up) and very advanced networking concepts not likely to be used in real-world systems for decades. Although they do profess support for IPv6, they have not aggressively deployed it, and it is definitely not central to their efforts. They are doing little or no work on IPv6 itself or in new commercial applications based on IPv6. I guess those areas are not very exciting to academicians. They are very exciting to me – actually, more exciting than 100 Gbps links.
The real-world Third Internet I am writing about in this book will be built primarily with equipment that mostly has the same performance as the current Second Internet (no more than 1 Gbps on WAN links for some time to come and only that high in advanced countries). In much of the world today, 5–120 Mbps is considered good. Maybe 100 Gbps will be widely deployed by 2030–2040, but ultrahigh performance is not necessary to provide the revolutionary benefits described in this book. To give you an idea, Standard-Definition (SD) TV requires about 2 Mbps bandwidth per simultaneously viewed channel, and High-Definition (HD) TV requires about 10 Mbps bandwidth. That is about the most bandwidth-intensive application you will likely see for most users for some time to come. Voice only requires about 8–64 Kbps for good quality. In Japan and Korea today, home Internet accounts typically have about 50–100 Mbps performance. In my hotel room in Tokyo several years ago, I measured 42 Mbps throughput. That is enough for almost any use today. I now have 1 Gbps Internet service in my home in SG (for about S$49 a month). Most users, even in companies, would be really challenged to make effective use of 100 Gbps bandwidth. With that bandwidth you could download the entire Encyclopedia Britannica in just a few seconds (including images) or a typical Blu-ray movie (about 25 gigabytes) in about 2 seconds. With current caps on network traffic volume, you would go through your entire month’s allowance in a matter of seconds. That is actually a serious concern even with 5G, with 1 Gbps potential speeds.
The necessary equipment and applications for the Third Internet can in many cases be created with software or firmware upgrades (except for older and low-end devices that don’t have enough RAM or ROM to handle the more complex software and in high-end telco- and Internet Service Provider (ISP) –level products that include hardware acceleration).
The main technical advantages of the Third Internet will not be higher bandwidth, but the vastly larger address space, the restoration of the flat address space (elimination of NAT), and the general availability of working multicast. All these are made possible by migration to IPv6, which involves insignificant costs compared with supporting 100 Gbps WAN links. Perhaps generally available WAN bandwidth in that range will be what characterizes the Fourth Internet. I personally would just consider that “faster Third Internet.”
Illustrations of two logos. The first is for Internet 2 with the number 2 at the center. The other one is of National LambdaRail with the letters N L R at the top.
Logos for Internet2 and National LambdaRail
Is It Web 2.0?
First, if you think that the terms “World Wide Web” and “Internet” are synonymous, let me expand your worldview a bit, in the same way that Copernicus did for people’s view of our Solar System back in the mid-1500s. The “World Wide Web ” is basically one service that runs on a much larger, more complex thing, which is called the Internet. The Web is a simple client-server system based on HTTP 62 (Hypertext Transfer Protocol ) and HTML 63 (Hypertext Markup Language ). Due to extremely serious limitations and inefficiencies of these standards, both have been enhanced and extended numerous times. The result is still not particularly elegant to real network software designers or engineers, but it has clearly had a major impact on the world. The technology of the Web was a refinement and convergence of several ideas and technologies that were in widespread use before HTML and HTTP were created by Tim Berners-Lee in the late 1980s, at CERN. But there is a lot to the Internet beyond the Web (email, instant messaging, video conferencing, VoIP, file transfer, peer-to-peer (P2P), VPNs, IPTV, etc.). There are thousands of Internet protocols, of which the Web uses one (HTTP).
Hypertext, WAIS/SGML, and Gopher
The terms Hypertext and Hypermedia were coined by Ted Nelson in 1965, at Brown University. These terms referred to online text documents (or rich media, including pictures, sound, and other media content) that contained links that allowed building paths from any word or phrase in the document to other parts of the same document or parts of other documents that were also online. In August 1987, Apple Computer released the first commercial Hypertext-based application, called HyperCard, for the Macintosh. There were already document storage and retrieval systems on the early Internet, such as WAIS 64 (Wide Area Information Server ). WAIS was based on the ANSI Z39.50:1988 standard and was developed in the late 1980s by a group of companies including Thinking Machines, Apple Computer, Dow Jones, and KPMG Peat Marwick. As with the Web, there were both WAIS servers and clients. A later version of WAIS was based on ANSI Z39.50:1992, which included SGML (Standard Generalized Markup Language , ISO 8879:1986) for more professional-looking documents. There was another Internet application called Gopher 65 (University of Minnesota, circa 1991) that could distribute, search for, and retrieve documents. Gopher was also primarily text based and imposed a very strict hierarchical structure on information.
HTML and HTTP
Tim Berners-Lee combined these three concepts (Hypertext, WAIS/SGML, and Gopher document retrieval) to create HTTP and HTML . HTML was a very watered-down and limited markup language compared with SGML. SGML is capable of creating highly sophisticated, professional-looking books. In comparison, HTML allows very limited control over the final appearance of the document on the client’s screen. HTTP was a very simple protocol designed to serve HTML documents to HTTP client programs called web browsers . A basic HTTP server can be written in one afternoon and consists of about half a page of the C programming language (I’ve done it and retrieved documents from it with a standard browser). The first browser (Lynx,66 1992) was very limited (text only, but including Hypertext links). In 1993, at the National Center for Supercomputing Applications (NCSA) at the University of Illinois, the first Mosaic 67 web browser was created (running on X Windows in UNIX). Because it was created for use on X Windows (a platform with good support for computer graphics), many graphics capabilities were added. With the release of web browsers for PC and Macintosh, the number of servers went from 500 in 1993 to 10,000 in 1994. The World Wide Web has since grown to millions of servers and many versions of the web client (Internet Explorer, Mozilla Firefox, Safari, Opera, Chrome, etc.). It’s been so successful that a lot of people today think that the “World Wide Web” is the Internet. It’s really just one small part of it.
Web 2.0
The term Web 2.0 68 was first coined by Darcy DiNucci in 1999, in a magazine article. The current usage dates from an annual conference that began in 2004, called “Web 2.0 ,” organized and run by Tim O’Reilly (owner of O’Reilly Media, publisher of many excellent books on computing).
Many of the promoters of the term Web 2.0 characterize what came before (which they call Web 1.0) as being “Web as Information Source.” Web 1.0 is based on technologies such as PHP, Ruby, ColdFusion, Perl, Python, and ASP (Active Server Pages). In comparison, Web 2.0 is “Network as Platform,” or the “participatory Web.” It uses the technologies of Web 1.0, plus new things such as Asynchronous JavaScript, XML, Ajax, Adobe Flash, and Adobe Flex. Typical Web 2.0 applications are the Wiki 69 (and the world’s biggest wiki, the Wikipedia 70), blogging sites, social networking sites like Facebook, video publishing sites like YouTube, photographic snapshot publishing sites like Flickr, Google Maps, etc.
Andrew Keen (British-American entrepreneur and author) claims that Web 2.0 has created a cult of digital narcissism and amateurism, which undermines the very notion of expertise. It allows anyone anywhere to share their own opinions and content, regardless of their talent, knowledge, credentials, or bias. It is “creating an endless digital forest of mediocrity: uninformed political commentary, unseemly home videos, embarrassingly amateurish music, unreadable poems, essays and novels.” He also says that Wikipedia is full of “mistakes, half-truths and misunderstandings.” Perhaps Web 2.0 has made it too easy for the mass public to participate. Tim Berners-Lee’s take on Web 2.0 is that this is just a “piece of jargon.” In the finest tradition of Web 2.0, these comments, which were found in the Wikipedia article on Web 2.0, probably include some mistakes, half-truths, and misunderstandings.
Basically, Web 2.0 does not introduce any revolutionary new technology or protocols; it is more a minor refinement of what was already being done on the Web, in combination with a new emphasis on end users becoming not just passive consumers, but also producers of web content. The Third Internet will actually help make Web 2.0 work better, as it removes the barriers that have existed in the Second Internet since the introduction of NAT to anyone becoming a producer of content. If anything, on the Third Internet, these trends will be taken even further by decentralizing things. There will be no need for centralized sites like YouTube or Flickr to publish your content, just more sophisticated search engines or directories that will allow people to locate content that will be scattered all over the world. Perhaps that will be the characterizing feature of Web 3.0? With IPv6 you can run any server (including a web server) on any computer you have, including your phone, and anyone in the world (who has IPv6) will be able to access it. Now that’s a major change.
A logo of web 2.0. Web 2 point 0 is written in the center of the logo and the services provided by it are written behind it against a dark background.
Web 2.0 logo
Whatever Happened to IPv5?
Two of the common questions people ask when they start learning about IPv6 are “If it’s the next version after IPv4, why isn’t it called IPv5 ?” and “What happened to the first three versions of IP?”
There is a 4-bit field in every IP packet header that contains the IP version number in binary. In IPv4, that field contains the binary value 0100 (4 in decimal) in every packet. An earlier protocol (defined in RFC 1190,71 “Experimental Internet Stream Protocol, Version 2 (ST-II),” October 1990) used the binary pattern 0101 (5 in decimal) in the IP Version field of the packet header. The Internet Stream Protocol was not really a replacement for IPv4 and isn’t even used today, but unfortunately the binary pattern 0101 was allocated to it. The next available bit pattern was 0110 binary (6 in decimal). It would be even more embarrassing than explaining that there was no IPv5 to explain why the IP version number field for IPv5 contained the value 6. Now you know.
So what did happen to IPv1, IPv2, and IPv3? Those never made it out of the lab. The first version of IP that was released to the general public was IPv4.
ARPANET 72 (based on NCP) was the First Internet. It didn’t use any version of the Internet Protocol – it used NCP. IPv4, the foundation protocol of the Second Internet, was the first public release of the Internet Protocol. IPv6, the foundation protocol of the Third Internet, is the second public release of the Internet Protocol. So we could have been talking about the transition from IPv1 to IPv2!
There have been rumors about an IPv9 73 protocol in China. A venture capital firm in Hong Kong actually asked me if China was already that far ahead of the rest of the world, and shouldn’t we be supporting their version? It seems some researcher in a university there published a paper on an “IPv9,” but it was never implemented and wasn’t a replacement for IPv4 (let alone IPv6) anyway. It was a way to use ten-digit decimal phone numbers in a modified DNS implementation instead of alphanumeric domain names, for all nodes on the Internet. I guess if you speak only Chinese, a ten-digit numeric string may seem easier to use than an English domain name using Latin characters. Fortunately for Chinese speakers, we now have Internationalized Domain Names 74 in Chinese and other languages.
There are even internationalized top-level domains (TLDs) now. For an example, see https://www.101domain.com/%E4%B8%AD%E5%9B%BD.htm .75
Actually, there is a real RFC about IPv9, which you might enjoy reading. See RFC 1606,76 “A Historical Perspective on the Usage of IP Version 9,” April 1, 1994. This has nothing to do with the Chinese IPv9 and is much funnier. Please notice the release date of RFC 1606. There is a tradition of releasing gag RFCs on April 1. Some of them are hilarious (well, maybe you have to be a geek to see the humor).
Let’s Eliminate the Middleman
One of the things that the Third Internet does better than anything is disintermediation .77 Just as email eliminated the need for a central post office and Amazon.com has mostly eliminated the need for physical bookstores, the features of the Third Internet will eliminate the need for many other existing centralized organizations and services. With a real decentralized end-to-end connectivity model, there is no need for two users to connect to a central server (such as Skype or Messenger) in order to chat with each other. They will simply connect directly to each other. That’s hard to do today, because of NAT and an acute shortage of public IPv4 addresses.
The restoration of the original (pre-NAT) flat address space and the plethora of public addresses will allow anyone or anything to connect directly to anyone or anything on the Third Internet. It’s going to be a very different online world. Many business models will go by the way, and many new ones will explode on the scene and make some new entrepreneurs very wealthy. Someone will need to provide centralized directory and presence servers that will let people locate each other, so that they can connect directly to each other. I am working on that very problem now.
Several years ago, a gentleman in my previous hometown of Atlanta, Georgia (home to Coca-Cola and UPS), had a small UHF TV station (WTBS, channel 17) that mostly broadcast old movies and Atlanta Braves baseball games, both of which he loved. He was one of the first people to realize that he could relay his TV station’s signal through a transponder on a geostationary satellite (“that’s just a really tall broadcast antenna”), and the rest is history. The man was Ted Turner,78 and his insight created the Turner Broadcasting System (TBS) , which along the way produced CNN, CNN Headline News, Cartoon Network, Turner Network Television (TNT), and many other things. His success allowed him to buy the Braves baseball team and the entire film library of MGM (not to mention a famous starlet wife, sometimes also called “Hanoi Jane”). When he began relaying his channel 17 signal, his viewership went from maybe 10,000 to 10,000,000 virtually overnight. That was a world-changing insight.
Some bright entrepreneur is going to realize that global multicast IPTV over IPv6 is the same kind of opportunity. Wonder what they will create with the wealth thereby generated? What country will they be from? I’m betting on India.
I did warn you that this is revolutionary, highly disruptive technology. However, with great disruption comes great opportunity.
Why Am I the One Writing This Book? Just Who Do I Think I Am, Anyway?
I have been personally involved in helping create and deploy the Third Internet for many years. I’ve spoken at IPv6 summits around the world, including Beijing, Seoul, Kuala Lumpur, Manila, Taipei, Potsdam, and Washington, DC. I have so far invested 25 years of my life and about $9M of my own personal funds (which came from selling a previous Internet-based venture called CipherTrust where I was cofounder). I’ve built a new company in Singapore with lots of expertise in PKI and IPv6. It is called Sixscape Communications 79 – we are “The Netscape for the IPv6 Internet.” The Third Internet is by far the biggest business and technology opportunity I’ve seen in my 45+ years in IT.
Now, have I gotten your attention? Great! Now let’s explore just what the Third Internet is all about.
Summary
This chapter covered the three generations of the global Internet, from ARPANET to IPv4 to IPv6. IPv6 is not just another version of one of the many Application Layer protocols like HTTP v1.1. It is deep in the network stack (in the Internet Layer), so it affects all Application Layer protocols.
The First Internet (ARPANET) only served a few thousand people, mostly in the United States. Most of the users were in the US military, the US government, and a few research institutions (mostly universities). It had 8-bit addresses and was based on the host-host protocol. It had many of the applications we still use today, like email, FTP, chat, etc. It lived from roughly 1969 to 1982.
The Second Internet (IPv4 generation) took over from the ARPANET in 1983 and grew to serve billions of users, worldwide. It used 32-bit addresses, but we ran out of unique IPv4 public addresses in 2011 and “broke” the Internet with NAT and private addresses. Many people are still using this today, but it will eventually be phased out.
The Third Internet (IPv6 generation) began major deployment in about 2014 and will probably still be in widespread use in 2100. It uses 128-bit addresses, which means we no longer need NAT or private addresses. Every node can have a globally unique IP address, even phones and temperature sensors.
We also presented some analogies to help you understand what IPv6 is and what it isn’t.