Internet Basics - A Technical Overview
Updated: 10th June 2010
The Internet started out as a research project over 40 years ago. It's hard to believe it's actually that old, but despite changing names and having many facelifts, it's still the same Internet underneath.
A Little History
The Internet was first known as ARPANET - the Advanced Research Projects Agency, and although it had military ties, its main purpose was research. ARPANET was nothing like what the Internet is today, there were no webpage's, no blogs, not even any email - the most it could do was access remote mainframes locked away in Universities and government establishments using custom build programs, that were evolving almost daily. These customized programs were designed to communicate with other 'nodes' on ARPANET, however almost every device was different in design - creating the first big hurdle towards global communication. Somewhere along the line, things had to be standardized.
The ARPANET as it looked in March, 1977
This challenge took up most of the 1970's for programmers working on the ARPANET or attending UCLA or Berkeley at the time, however it wasn't long after that the first signs of TCP/IP (Transmission Control Protocol/Internet Protocol) came about - the protocol that is still used on the Internet to route traffic. The short version of the story is that the two main host-to-host protocols NCP and IMP were replaced with TCP/IP around 1984, the rest is history.
However, what we have now hasn't changed all that much since the late 1980's. Granted the Internet is (arguably) faster, and pages look nicer, the underlying technology is still the same as it has been for a long time. The benefit to this is of course learning about the web is far easier, as there is an abundance of information available, and the technology to implement it is now relatively cheap. But the downside is that the current Internet implementation has struggled to keep up with demand, and won't last long. We're soon due for a new net revolution.
Note:
I thought I'd choose now to clear up my position on this Internet doomsday theory that is very quickly becoming rumour. Yes it's true the world is fast running out of IPv4 addresses; however that DOES NOT mean the Internet will suddenly disappear. Over the coming years, I'm sure the media will portray the declining amount of free IP addresses as the next Millennium bug or something, but you now know better…, and the next-generation solution (IPv6) is already being rolled out in some parts of the world.
The Physical Layer
Now you have a little knowledge about where the Internet came from, and where it's going… it's time to unravel what it actually is. Feel free to check the right hand box for a glossary of terms commonly used on the Internet.
Chances are if you're reading this, somewhere, someone has connected you to the Internet. However the Internet is a rather ambiguous term, it's not something you can physically touch, or see. You can't simply turn off the Internet or treat it as a whole entity anymore… it's simply too big. It's now just a generic term for all the network devices that share a public network to access each other.
It's also a collection of information, databases, users and groups that make up the various societies of the Internet, but any social expert can break it down into these terms, we're interested in the nuts and bolts.
The Internet is a public network. Meaning it crosses paths which almost anyone in the world can use. These generally are national and international phone networks, however since the mid 90's; Telco's have installed dedicated data lines across oceans and continents purely to aid the growth of the Internet. Because it spans so many countries and borders - no one owns the Internet, it's simply a name.
That's not to say no one is in control of it. There are various groups in place who oversee the health and welfare of this massive network, IANA (Internet Assigned Numbers Authority) is one such organisation who ensure the Internet protocols function correctly, and the ISO (International Organization for Standardization) ensure all devices that want to communicate can do so by speaking the same language (among other things). If it wasn't for these organisations, the Internet would only be half of what it is today, and it certainly wouldn't be available to anyone.
The ones with the real power though are the ISP's (Internet Service Providers). They control the connections between their networks and others - and that's all the Internet really is; a vast internetwork. ISP's own and operate the public telephone networks which the Internet started piggy-backing on in the late 80's, so it made sense that they would eventually become the soul- beneficiaries of the physical cabling that makes up the Internet. It also came with a lot of responsibility, something which ISP's are slowly adjusting to. In short, the ISP's of the world 'inherited' the Internet through their existing physical infrastructure, and today – data exceeds voice, with more and more of the backbone systems converting voice calls to data, allowing both phone conversations and Internet signals to transfer over the same physical cabling.
What are these signals, and where is this content?
Almost all web traffic (not including spam and viruses!) is web traffic. Event applications like Skype (that use VOIP, or voice over IP) rely on data packets over the web. Every page you visit is a file on another machine, and you are simply requesting to view that file, For example, the text your reading now is stored in a file on The Serpent which is in turn hosted by 1and1 Internet. Physically 1and1 Internet has a server (or mainframe) with disks containing this file and your browser simply downloads, interprets and handles the displaying of that file. It all sounds so simple, and it really is on the surface, the true power of the Internet shines when you request a file that's physically located thousands of miles away. Retrieving a small HTML file suddenly becomes a complex process of routing, multiplexing and transformations.
A to B, Via C, D, E, F...
Internet routing is the topic of some rather thick books, so after skimming the surface, I'll let you decide if you'd like to know more. In order for the Internet to span the distance it does, corporation from many different companies is required. It's less of an issue nowadays then it was ten years ago, but in order to reach far out lands, phone companies need to make deals with each other so that data can be routed properly.
Generally speaking, each ISP is known as an autonomous system - an entity that deals with its own internal structure and only advertises to the world the services it has. To make this simple - each device, weather that is a server or client needs its own unique address that is usable on the Internet. ISP's advertise these IP (Internet Protocol) addresses that they host to their neighbouring ISP's, creating a mesh of routes which allow data to traverse the planet. Since the choice of communication is TCP/IP - IP addresses form the basis for any interconnected device. As mentioned previously, IP's are a finite resource, and must be regulated properly to ensure they remain unique across the globe – not an easy challenge when there are a total of 4,294,967,296IP addresses within the system!
The job of allocating these IP's fall on the Internet Assigned Numbers Authority, an organisation dedicated to ensuring we get the most out of our IP's, amongst their responsibilities, they ensure ISP's have blocks of IP's which don't overlap. Each ISP is assigned their own unique block, and they use these addresses in everything from inter-connections to other ISP's, servers hosting content to customer's broadband connections.
The IP Address
When you connect up your broadband modem or wireless dongle on your Laptop, you're given an IP address from your ISP. This IP uniquely identifies you from all the other customers on the same ISP. It also tells other ISP's which ISP you belong to, so should you request something from the other side of the world, all the intermediate devices know to send the traffic to the same ISP – yours.
In reality it's a little more involved than that, but it's a rough overview. There are situations where IP's can be duplicated (such as private LAN IP's on totally separate LAN's), but the general rule is that your Internet IP is your own; no other Internet connected device has it. This can be a good thing, and a bad thing.
Note:
There's a lot of software out there that claims to 'mask' your IP address, but in reality, does nothing more than a free firewall package would do. It's analogous to masking your phone number and wondering why no one can call you. Your IP address needs to be known by remote hosts if you want them to transfer data to you, even if you initiated the request, so it's generally a bad idea to give into software which claims to mask your Internet identity.
DNS – The Internet Phonebook
Since the Internet relies on IP addresses, and hosts never use names to talk to each other, how does your machine know the IP address of some Google server far far away? Simple – we use DNS.
DNS (Domain Name System) is the Internet phonebook; it's a database which maps easy to remember hostnames such as 'www.google.com' to an IP address that ISP's can physically locate. All main ISP's maintain a partial copy of the DNS database, but it's centrally managed via a group of root servers across the globe, allowing for redundancy, speed and security within the system.
Your ISP provides you with the IP address of a DNS server which gives you access to the database, before your browser attempts to locate www.google.com; it sends the name to your defined DNS server, which in turn returns the IP you need. Your computer then continues the request with one of Google's many IP's, and you don't even know it.
DNS isn't without its flaws, when the DNS server is unavailable to perform the lookup, you'll probably get a common 'unresolved hostname' error or something similar, most ISP's provide you with two DNS server addresses, so you can try each in turn, to provide more stability, most large sites (such as Google) resolve to multiple IP's instead of just one, so a single name could map to hundreds of devices, allowing servers to spread across entire farms.
You can see DNS in action by bypassing it all together, type the following into your browser to get to Google without even using its name:
http://66.102.7.99/
Since your computer actually prefers numbers, this simply tells the computer to skip the DNS lookup and head to one of Google's IP's.
The Internet vs. WWW
Contrary to popular belief, the Internet is not the same thing as the World Wide Web. The main difference being that the WWW encompasses the web site aspect, a collection of pages distributed using HTTP via the Internet (as opposed to other Internet services that don't rely on WWW, such as email, IRC, Torrents etc). The Serpent is on the WWW, and distributed via the Internet, but as times change, and since the WWW is by far the most widely used aspect of the Internet, it's become natural to refer to the WWW as the Internet.
The Internet can get a lot more technical, and this is about as basic as it gets. Check out the rest of the site for articles which dig deeper into the individual technologies mentioned above, and see how far you're willing to go!
