A short history of the internet for networking students
The internet stands for many things: an impressive testament to the value of government-aided technological progress; harbinger of 21st-century “design by committee” innovation; the new “electricity”; the technology that heralded in the information age. Analogous to the modern age, it’s complicated and dynamic: a never to be completed, ever-moving target of innovation. It’s all of the above–and more, but pared down to its core it’s just computer networking: the simple act of computers exchanging information.
The internet’s development was a convoluted process, as its creation followed no definitive roadmap nor had one singular inventor. Rather it emanated from a cluster of smaller projects that had their roots in advancing telecommunications: the transmission of information by various types of technologies over wire, radio, optical, or other electromagnetic systems. These developments were tied to the invention of the transistor and the advent of and advance of computers. Simply put, now that we had powerful computers, the use case for connecting them together was compelling. Enterprising individuals and institutions started building transmission protocols, and just like that, the internet started to take shape, like a jigsaw puzzle that was ultimately put together under a central authority. Viewed historically, the internet can be as much a lesson in 21st techno-bureaucracy, as it is in the technical disciple of organizing 0s and 1s for transmission and reception.
Studying an immense topic like the internet can feel overwhelming. After all, it’s a technology that has expanded to serve 5.07 billion people daily—that’s 63.5% of the population! Choosing to study its history first is a good idea as it serves as an interesting point departure that also includes foundational technical knowledge. Below will be a cursory review of the internet’s history, relating it to the present with the aim of bringing perspective to this modern technology we take advantage of–and for granted, every day.
After the historical review, I will detail the construction and implementation of an small business Windows Hyper-V solution for ten end users, ultimately running several networking-centric enterprise technologies:
Routing Hardware: Mikrotik Router and Synology Wi-Fi
Virtualization Server: a tower PC build with Supermicro MB + NVME + RAID5. With Windows Server 2019 with the Hyper-V addon.
Stateful Firewall: pfSense based Software Defined Routing with an Intel NIC card + old PC + pfSense
Enterprise networking tools: pfSense addons: pfBlocker (firewall), Tailscale (VPN), Suricata (IDS/IPS) and an OrangePi running Security Onion
…with me reflecting along the way with lessons learned, to frame the process as an educational exercise.
What is having the internet?
If you want to listen to the transmission of sound, there needs to be air, to carry the sound energy, and an ear or otherwise device to “hear” the noise. A simple form of noise creations, banging two things together, creates a mechanical wave. Armed with this information, we can give an obtuse answer to the parable:
"If a tree falls in a forest and no one is around to hear it, does it make a sound?"
1) Yes, a sound will be made as the mechanical wave propagates itself on nearby mediums like air or tree bark reverberating, etc.
2) Yes/No, as it will only be heard if a device is present to detect and transfer the mechanical wave energy back to noise or information.
Drawing inspiration from these natural systems, humans extracted and applied mathematical theorems to create and enhance communication methods. This began with the use of metal telephone wires, exemplified by Alexander Graham Bell's invention of the telegraph in 1874, and later evolved to encompass wireless electromagnetic wave propagation via antennas, as pioneered by Guglielmo Marconi in 1895.
Whether through air or wire, we need a medium to send signals and a corresponding receiver to catch and decode them. With a radio, the signal receiver is attached to a speaker to translate that signal back into sound, although it can now be wholly saved digitally for playback later. For television, we do much the same, somehow receiving a compound signal of audio and video information through radio waves or wire–often coaxial copper cabling. To help disperse these signals, a network of sorts was created for television. An early implementation of this was Community Antenna Television (CATV), which consisted of a community antenna that would receive a broadcast signal and then rebroadcast it locally, leading to today’s network affiliates.
Phones similarly rely on a networking of cables that enable their messages to be distributed and rebroadcast, as they travel to their destinations. Seen in context with these technologies, the internet can be seen as a natural progression of these early information exchange networks. And just as cable TV utilized existing phone infrastructure, the internet utilizes both previous infrastructures today: coaxial cable and the original balanced pair phone line.
All the above technologies were and are technologies whose central role was to distribute any data that could be recorded. This process is akin to quantification: in mathematics and empirical science, quantification (or quantitation) is the act of counting and measuring that maps human sense observations and experiences into quantities. This quantification process has enabled us to collect and share data as never before.
For instance, just think how constrained one was to place and time, as say, if you wanted to attend a play. You would need a ticket to reserve a physical space at a specific, unique time for you to experience the production as it occurs on stage. With quantification (e.g. video recording), that experience can mostly be recreated on demand, freeing it from the strictures of time and place. Thinking along these lines, the first phone call below can be seen as an early version of an internet chatroom.
Getting on the map, an ‘internet address’
What can we do to buy a ticket to the internet? One might start with an “access point” that is a unique 32-bit binary IPv4 or 128-bit hexadecimal Ipv6 address leased to you by an internet service provider (ISP). This will act as your name on the network, and you will use it as a home address so that you can send and receive messages from its “location” as a node belonging to the network. The rules and system that regulates the dispensation of the addresses is regulated by the IETF and is called the Domain Name System. Your ISP will be allowed use of a range of addresses and their customers are leased an address. Ultimately, a network interface card (NIC) or access point router will take on this address to be used by your systems on the public internet. There is a distinction that you can use whatever addresses you wish in your private network, but on the public internet, you must have a unique IP. Being a core item to being on the network, this IP address can be found on the L3 “network layer.”
Layer one of what? Well, the internet is often summarized abstractly as a discretely separated, but interweaving, layered model. As you will see, the layers pass data to one another by packaging in with headers and footers. Regarding the addressing system, a user actually will have two addresses, one physical for layer one (L1) and the other abstracted according to the domain name system that was implemented for internet protocol (IPv4 or IPv6). The IP address exists above layer one on layer two (the data link layer) and is a necessary component to operate as a host node on the network.
Early implementations of packet-switching protocols
The above mentioned IP address is part of the Internet protocol suite (TCP/IP Model (Transmission Control Protocol/ Internet Protocol), the core component of the modern internet. It was initially developed by the Defense Advanced Research Projects Agency (DARPA) in the late 1960s, and is still now the utilized protocol. In your studies, you will find that these protocols are responsible for ordering, verifying, and transmitting data. But before that discussion, let us review how the TCP/IP standard came to fruition.
Today, the technical standards underlying the internet protocol suite are maintained by a division of the United Nations: the Internet Engineering Task Force (IETF).
In addition to being familiar with this, one should know of the OSI model, a conceptual model: a comprehensive design document for building networking systems. Reading internet history, you will find that many models have helped direct the internet’s evolution by giving structure to the myriad of different protocols and packet switching technologies that were devised in the 1970s and 80s. This period of history is known as the protocol wars.
To get a sense of the net’s lineage, one can look to the The International Telecommunication Union (ITU) another specialized United Nations agency for information and communication technologies that is the predecessor of the IETF. It was founded in 1865 by Napoleon III to direct the standardization of the telegraph. They have since published over 4000 recommendations. One of their contributions related to the internet is the European based standard X.25:
(being published in 1976) …it is one of the oldest packet-switching communication protocols available; it was developed several years before IPv4 (1981) and the OSI Reference Model (1984). The protocol suite is designed as three conceptual layers, which correspond closely to the lower three layers of the seven-layer OSI model. It also supports functionality not found in the OSI network layer.
That being said, for practical purposes, it’s sufficient today to just understand the OSI model and the TCP/IP protocol, as well as the many improvements that have been added by IETF and vendors. The internet’s history is messy, a side effect of its collaborative, iterative nature that has its roots in the academic and institutional R&D domains.
OSI and TCP/IP to send PDUs
As a networking engineer, you should become familiar with the OSI model. Several have been formulated, but for this blog, I will exclusively reference the International Organization for Standardization’s (ISO) 7-layer model that was published in 1984. Starting with layer 1 it goes: Physical, Data Link, Network, Transport, Session, Presentation, and Application.
Each layer groups a number of technologies that perform various functions in the network.
Referencing the above model, your public IP address (AKA IP) exists on layer 3, the Network layer. Your IP will be utilized to identify yourself and other computers in network. In essence, you are purchasing access to the internet through an ISP that acts as a courier service to send and receive data for you through the internetworked mesh that is the ‘internet.’ Further discussion on ISP particulars is not pertinent at this moment, suffice to say that telecommunications companies run high-capacity cables to coordinate web traffic between hosts and servers, utilizing a specialized L3 routing protocol called Border Gateway Protocol (BGP). A tertiary concept you may have heard about is “edge computing.” Per Wikipedia, “The origins of edge computing lie in content distributed networks that were created in the late 1990s to serve web and video content from edge servers that are deployed closer to users for quick access.
Returning to our public IPv4 address, we may realize that one address is not enough for the multiple devices that exist in our household. Remember, public IPs must be unique. Thus, our address must be shared somehow, to allow for multiple users to still utilize one address while keeping their traffic separate. A hardware device called a router can be introduced to solve this problem through a process called Port Address Translation (PAT).
As an aside, in networking there are often multiple types of the same thing. I think of this as a reminder of the duplicative nature of the system. For instance, data has many names or forms based on its context and position in the system. Each layer of the model has its own protocol data unit (PDU): L1 Bits; L2 Frames; L3 Packets; L4 Segments. As you learn about how the network functions, you will understand how the data continually modified, packaged and re-packaged to coordinate its travel along the network path.
The two most central forms of this re-packaging occur at two different localities of the internet, L2 and L3. For L2, remember Data-link is always network local using MAC addressing. If data needs to leave the local network, it will need an IP address to operate on L3. L2 is much simpler, and operates with a few additional technologies to enable a frames movement from within a local network. Unlike IP addresses, which do not change at each hop, L2 frames have their source and destination address replaced to direct their movement at each router (AKA frame rewrite. Switches do not perform frame rewrites; they exclusively pass frames along per their Content Addressable Memory (CAM) and tagged VLAN information.
For out of LAN traffic (AKA inter-network traffic), L3 Network Layer implements routing algorithms that act as a ‘traffic cop protocol,’ directing packets in the most efficient manner possible. An important point to be made is that packets are not self-determining, ie, they merely contain their data payload inside L3 with additional meta-data in the header/footer that is utilized by whichever L3 technology it encounters. Packets can pass through various protocols before reaching their destination.
To summarize, the TCP/IP protocol is only one of many L3 “Network” protocols that go about executing OSI’s dictum: the relaying of packets from one network segment to another by nodes in a computer network via packet forwarding. Including routing packets through intermediate routers.. A network trace route command can be used to read back information concerning the packet’s travel, e.g., how many “hops” (trips between routers) a packet has made.
These “Network” protocols are algorithmic logic that are usually run-on ASIC hardware embedded, although software implementations exist. For example, three routing protocols are supported by the networking BSD OS pfSense. In the next post, we will plan and set up a home network utilizing a pfSense computer and a router.
References
- {/* */}↩