---
tags: Google IT support, Networking
---

# The Bits and Bytes of Computer Networking # 05 - Connecting to the Internet

## POTS and Dial-up

### Dial-up, Modems and Point-to-Point Protocols

In the late 1970s two graduate students at Duke University were trying to come up with a better way to connect computers at further distances. They wanted to share what was essentially bulletin board material, then a light bulb moment went off. They realized the basic infrastructure for this already existed, the public telephone network.

- **The public Switched Telephone Network (PSTN)**
The public Switched Telephone Network or PSTN is also sometimes referred to as **the Plain Old Telephone Service** or **POTS**. It was already a pretty global and powerful system by the late 1970s more than 100 years after the invention of the telephone. These Duke grad students weren't the first ones to think about using a phone line to transmit data. But they were the first do it in a way that became somewhat permanent precursor to the **dial up networks** to follow. The system they built is known as **USENET** and a form of it is still in use today.

A dial-up connection uses POTS for data transfer, and gets its name because the connection is established by actually dialing a phone number.

Transferring data across a dial-up connection is done through devices called **modems**.

- **Modems**
Stands for modulator demodulator, and they take data that computers can understand and turn them into audible wavelengths that can be transmitted over POTS. After all, the telephone system was developed to transmit voice messages or sounds from one place to another. This is conceptually similar to how line coding is used to turn ones and zeroes into modulating electrical charges across Ethernet cables. Early modems had very low **baud rates**.
- **Baud Rates**
A measurement of how many bits could be passed across a phone line in a second

By the late 1950s, computers could generally only send each other data across a phone line at about a 110 bits per second. By the time USENET was being developed, this rate had increased to around 300 bits per second. And by the time dial-up access to the Internet became a household commodity in the early 1990s, this rate had increased to 14.4 kilobits per second.

Dial-up Internet connectivity is pretty rare today but it hasn't completely gone away. In some rural areas, it might be the only option still available. You might never run into a dial-up Internet connection during your IT career. But it's still important to know that for several decades this technology represented the main way computers communicated with each other over long distances.

## Broadband Connections

In terms of internet connectivity, it's used to refer to any connectivity technology that **isn't dial-up** Internet. Broadband Internet is almost always much faster than even the fastest dial-up connections and refers to connections that are always on. This means that they're long lasting connections that don't need to be established with each use. They're essentially links that are always present.

Broadband shaped today's world. While the Internet itself is a totally amazing invention, it wasn't until the advent of broadband technologies that its true potential for business and home users was realized. Long before people had broadband connections at home, businesses spent a lot of resources on them usually out of necessity.

By the mid 1990's, it had become pretty common for businesses that needed internet access for their employees to use various **T-carrier technologies**.

- **T-carrier technologies**
Originally invented by AT&T in order to transmit multiple phone calls over a single link. Eventually, they also became common transmission systems to transfer data much faster than any dial-up connection could handle.

T-carrier technologies require dedicated lines, which makes them more expensive. For this reason, you usually only see them in use by businesses. But other broadband solutions also exist for both businesses and consumers.

### T-Carrier Technologies

T-carrier technologies were first invented by AT&T in order to provision a system that allowed lots of phone calls to travel across a single cable. Every individual phone call was made over individual pairs of copper wire before **Transmission System 1**, the **first T-carrier specification**, called **T1** for short.

With the T1 specification, AT&T invented a way to carry up to 24 simultaneous phone calls across a single piece of twisted pair copper. Years later, this same technology was repurposed for data transfers. Each of the 24 phone channels was capable of transmitting data at 64 kilobits per second, making a single T1 line capable of transmitting data at 1.544 megabits per second. Over the years, the phrase T1 has come to mean any twisted pair copper connection capable of speeds of 1.544 megabits per second, even if it doesn't strictly follow the original Transmission System 1 specification.

Originally, T1 technology was only used to connect different telecom company sites to each other and to connect these companies to other telecom companies. But with the rise of the Internet as a useful business tool in the 1990s, more and more businesses started to pay to have T1 lines installed at their offices to have faster Internet connectivity. More improvements to the T1 line were made by developing a way of multiple T1s to act as a single link. So a **T3** line is **28 T1s**, all multiplexed, achieving a total throughput speed of 44.736 megabits per second.

You'll still find T-carrier technologies in use today, but they've usually been surpassed by other broadband technologies. For small business offices, cable broadband or fiber connections are now way more common since they're much cheaper to operate. For inter-ISP communications, different fiber technologies have all replaced older copper-based ones.

### Digital Subscriber Lines

For a long time, dial-up connections were the main way that people connected to the Internet from home. But there were certain limitations with trying to transmit data as what were essentially just **audio waves**. As people wanted faster and faster Internet access, telephone companies began to wonder if they could use the same infrastructure but in a different way. The research showed that twisted pair copper used by modern telephone lines was capable of transmitting way more data than what was needed for voice-to-voice calls. By operating at a frequency range that didn't interfere with normal phone calls, a technology known as **digital subscriber line** or **DSL** was able to send much more data across the wire than traditional dial-up technologies. To top it all off, this allowed for normal voice phone calls and data transfer to occur at the same time on the same line.

Like how dial-up uses modems, DSL technologies also use their own modems. But, more accurately, they're known as **DSLAMs** or **Digital Subscriber Line Access Multiplexers**.

Just like dial-up modems, these devices establish data connections across phone lines, but unlike dial-up connections, they're usually long-running. This means that the connection is generally established when the DSLAM is powered on and isn't torn down until the DSLAM is powered off. There are lots of different kinds of DSL available, but they all vary in a pretty minor way. For a long time, the two most common types of DSL were **ADSL** and **SDSL**.

- **ADSL**
Stands for **Asymmetric Digital Subscriber Line**. ADSL connections feature different speeds for outbound and incoming data. Generally, this means faster download speeds and slower upload speeds. Home users rarely need to upload as much data as they download since home users are mostly just clients.
- **SDSL**
Stands for Symmetric Digital Subscriber Line. SDSL technology is basically the same as ADSL except the upload and download speeds are the same. At one point, SDSL was mainly used by businesses that hosted servers that needed to send data to clients.

As the general bandwidth available on the Internet has expanded and as the cost of operation have come down over the years, SDSL is now more common for both businesses and home users. Most SDSL technologies and have an upper cap of 1.544 megabits a second or the same as a T1 line.

Further developments in SDSL technology have yielded things like **HDSL** or **High Bit-rate Digital Subscriber Lines**. These are DSL technologies that provision speeds above 1.544 megabits per second.

### Cable Broadband

By using frequencies that don't interfere with television broadcast, cable-based Internet access technologies were able to deliver high speed Internet access across these same cables. This is the technology that we refer to when we say **cable broadband**. One of the main differences in how cable Internet access works when compared to other broadband solutions is that cable is generally what's known as a **shared bandwidth technology**. With technologies like DSL or even dial up, the connection from your home or business goes directly to what's known as a **Central Office** or **CO**. A long time ago, the COs were actually offices staffed with telephone operators who used a switchboard to manually connect the caller with the callee. As technology improved, the COs became smaller pieces of automated hardware that handled these functions for the telephone companies, but the name stayed the same. Technologies that connect directly to a CO can guarantee a certain amount of bandwidth available over that connection since it's point to point. On the flip side of this, are cable Internet technologies, which employ a shared bandwidth model. With this model in place, many users share a certain amount of bandwidth until the transmissions reach the ISP's core network. This could be anywhere from a single city block to entire subdivisions in the suburbs. It just depends on how that area was originally wired for cables.

Today, most cable operators have tried to upgrade their networks to the point that end users might not always notice the shared bandwidth. But it's also still common to see cable Internet connections slow down during periods of heavy use. Like when lots of people in the same region are using their Internet connection at the same time. Cable Internet connections are usually managed by what's known as a **cable modem**.

- **Cable Modem**
A device that sits at the edge of a consumer's network and connects it to **the cable modem termination system**, or **CMTS**.
- **Cable modem termination system(CMTS)**
Connects lots of different cable connections to an ISP's core network

### Fiber Connections

The core of the Internet has long used fiber for its connections, both due to higher speeds and because fiber allows for transmission to travel much further without degradation of the signal. The absolute maximum distance an electrical signal can travel across a copper cable before it degrades too much and requires a repeater is thousands of feet, but certain implementations of fiber connections can travel many, many miles before signal degrades. Producing and laying fiber is a lot more expensive than using copper cables. So for a long time it was a technology you only saw in use by ISPs within their core networks, or maybe for use within data centers. But in recent years it's become popular to use fiber to deliver data closer and closer to the end user. Exactly how close to the end user can vary a ton across implementations, which is why the phrase **FTTX** was developed.

- **FTTX**
Stands for fiber to the X, where the X can be one of many things
- The first term you might hear is **FTTN**, which means fiber to the neighborhood. This means that fiber technologies are used to deliver data to a single physical cabinet that serves a certain amount of the population. From this cabinet, twisted pair copper or coax might be used for the last length of distance.
- The next version you might come across is **FTTB**. This stands for fiber to the building, fiber to the business, or even fiber to the basement, since this is generally where cables to buildings physically enter. FTTB is a setup where fiber technologies are used for data delivery to an individual building. After that, twisted pair copper is typically used to actually connect those inside of the building.
- A third version you might hear is **FTTH**, which stands for fiber to the home. This is used in instances where fiber is actually run to each individual residents in a neighborhood or apartment building.
- FTTH and FTTB may both also be referred to as **FTTP** fiber to the premises. Instead of a modem, the demarcation point for Fiber Technologies is known as **Optical Network Terminator** or **ONT**.
- **ONT**
Converts data from protocols, the fiber network can understand to those that more traditional twisted pair copper networks can understand.

## WANs

Unlike a LAN or a local area network, WAN stands for **wide area network**.
- **WAN**
Acts like a single network, but spans across multiple physical locations. WAN technologies usually require that you contract a link across the Internet with your ISP. This ISP handles sending your data from one site to the other. So it could be like all of your computers are in the same physical location.

A typical WAN setup has a few sections. Imagine one network of computers on one side of the country and another network of computers on the other. Each of those networks ends at a demarcation point, which is where the ISP's network takes over. The area between each demarcation point and the ISP's actual core network is called a **local loop**. This local loop would be something like a T-carrier line or a high speed optical connection to the provider's local regional office. From there, it would connect out to the ISP's core network and the Internet at large. WANs work by using a number of different protocols at the data link layer to transport your data from one site to another. In fact, these same protocols are what are sometimes at work at the core of the Internet itself, instead of are more familiar ethernet.

### Point-to-Point VPNs

Whann technologies are great for when you need to transport large amounts of data across lots of sites, because Whann technologies are built to be super fast. A business cable or DSL line might be way cheaper but it just can't handle the load required in some of these situations. But over the last few years, companies have been moving more and more of their internal services into the cloud.

With these types of cloud solutions in place, lots of businesses no longer require extreme high speed connections between their sites. This makes the expense of a Whann technology totally unnecessary. Instead, companies can use point-to-point VPNs to make sure that there are different sites can still communicate with each other. A **point-to-point VPN**, also called a **site-to-site VPN**, establishes a VPN tunnel between two sites. It's just that the VPN tunneling logic is handled by network devices at either side, so that users don't all have to establish their own connections.

## Wireless Networking

A way to network without wires.

The most common specifications for how wireless networking devices should communicate, are defined by the **IEEE 802.11 standards**. This set of specifications, also called the **802.11 family**, make up the set of technologies we call **WiFi**.
Wireless networking devices communicate with each other through **radiowaves**. Different 802.11 standards generally use the same basic protocol, but might operate at different **frequency bands**.

- **Frequency bands**
A certain section of the radio spectrum that's been agreed upon to be used for certain communications. In North America, FM radio transmissions operate between 88 and 108 megahertz. This specific frequency band is called the **FM broadcast band**.

WiFi networks operate on a few different frequency bands. Most commonly, the 2.4 gigahertz and 5 gigahertz bands. There are lots of 802.11 specifications including some that exist just experimentally or for testing. The most common specifications you might run into are **802.11b**, **802.11a**, **802.11g**, **802.11n**, and **802.11ac**.
Each newer version of the 802.11 specifications has generally seen some improvement, whether it's higher access speeds, or the ability for more devices to use the network simultaneously. In terms of our networking model, you should think of 802.11 protocols as defining how we operate at both the **physical** and the **data link** layers. An 802.11 frame has a number of fields:


- **Frame control field**
16 bits long, and contains a number of sub-fields that are used to describe how the frame itself should be processed. This includes things like what version of the 802.11 was used.
- **Duration field**
It specifies how long the total frame is. So, the receiver knows how long it should expect to have to listen to the transmission.
- **Four address fields**
The most common setup includes devices called **access points**.
- **Wireless access point**
A device that bridges the wireless and wired portions of a network.

A single wireless network might have lots of different access points to cover a large area. Devices on a wireless network will associate with a certain access point. This is usually the one they're physically closest to. But, it can also be determined by all sorts of other things like general signal strength, and wireless interference.
Associations isn't just important for the wireless device to talk to a specific access point, it also allows for incoming transmissions to the wireless device to be sent by the right access point. There are four address fields, because there needs to be room to indicate which wireless access point should be processing the frame.
1. **Source address address**
Represent the MAC address of the sending device
2. **Intended destination**
The intended destination on the network
3. **Receiving address**
Would be the MAC address of the access point that should receive the frame
4. **Transmitter address**
Would be the MAC address of whatever has just transmitted the frame

In lots of situations, the destination and receiver address might be the same. Usually, the source and transmitter addresses are also the same. But, depending on exactly how a specific wireless network has been architected, this won't always be the case. Sometimes, wireless access points will relay these frames from one another.
Since all addresses in an 802.11 frame are Mac addresses, each of those four fields is **6 bytes long**.
- **Sequence control field**
16 bits long and mainly contains a sequence number used to keep track of ordering the frames.
- **Data payload**
Gas all of the data of the protocols further up the stack.
- **Frame check sequence field**
Contains a checksum used for a cyclical redundancy check. Just like how ethernet does it.

### Wireless Network Configurations

There are a few main ways that a wireless network can be configured. There are **ad-hoc networks** where nodes all speak directly to each other. There are **wireless LANS** or **WLANS** where one or more access points act as a bridge between a wireless and a wired network. And there are **mesh networks** which are kind of a hybrid of the two.

- **Ad-hoc networks**
In an ad-hoc network, there isn't really any supporting network infrastructure. Every device involved with the network communicates with every other device within range and all nodes help pass along messages. Even though they're the most simple, ad-hoc networks aren't the most common type of wireless network, but they do have some practical applications. Some smartphones can establish ad-hoc networks with other smartphones in the area so that people can exchange photos, video or contact information. You'll also sometimes see ad-hoc networks used in industrial or warehouse settings, where individual pieces of equipment might need to communicate with each other, but not with anything else. Finally, ad-hoc networks can be powerful tools during disaster situations. If a natural disaster like an earthquake or hurricane knocks out all of the existing infrastructure in an area, disaster relief professionals can use an ad-hoc network to communicate with each other while they perform search and rescue efforts.
- **Wireless LAN(WLAN)**
The most common type of wireless network you'll run into in the business world. A wireless LAN consists of one or more access points, which act as bridges between the wireless and wired networks. The wired network operates as a normal LAN, the wired LAN contains the outbound internet link. In order to access resources outside of the WLAN, wireless devices would communicate with access points. They then forward traffic along to the gateway router, where everything proceeds like normal.
- **Mesh networks**
Kind of like ad-hoc networks, since lots of the devices communicate with each other wirelessly, forming a mesh. If you were to draw lines for all the links between all the nodes, most mesh networks you'll run into are made up of only wireless access points and will still be connected to a wired network. This kind of network let's you deploy more access points to the mesh without having to run a cable to each of them. With this kind of setup, you can really increase the performance and range of a wireless network.

### Wireless Channels

- **Channel**
Individual, smaller sections of the overall frequency band used by a wireless network. Channels are super important because they help address a very old networking concern, **collision domains**.
- **Collision domain**
Any one network segment where one computer can interrupt another.

Communications that overlap each other can't be properly understood by the receiving end. So when two or more transmissions occur at the same time, also called a collision, all devices in question have to stop their transmissions. They wait a random amount of time and try again when things quiet down. The problem caused by collision domains has been mostly reduced on wired networks through devices called switches.
Wireless networking doesn't have cables, so there aren't physical interfaces for a wireless device to connect to. That means, we can have something that works like a wireless switch. Wireless devices are doomed to talk over each other. Channels help fix this problem to a certain extent.
For wireless networks that operate on a 2.4 Gigahertz band, what we really mean is that they operate on roughly the band from 2.4 Gigahertz to 2.5 Gigahertz. Between these two frequencies are a number of channels, each with a width of a certain megahertz. Since different countries and regions have different regulatory committees for what radio frequencies might be used for what, exactly how many channels are available for use depends on where in the world you are. For example, dealing with an 802.11b network, channel one operates at 2412 megahertz, but since the channel width is 22 megahertz, the signal really lives on the frequencies between 2401 megahertz and 2423 megahertz. This is because radio waves are imprecise things. So, you need some buffer around what exact frequencies a transmission might actually arrive on. Some channels overlap but some are far enough apart so they won't interfere with each other at all. Let's look again at 802.11b network running on a 2.4 Gigahertz band, because it's really the simplest and the concepts translate to all other 802.11 specifications. With a channel width of 22 megahertz, channel one with its midpoint at 2412 megahertz, is always completely isolated from channel six with its midpoint at 2437 megahertz. For an 802.11b network, this means that channels one and six and 11 are the only ones that never overlap at all.
Today, most wireless networking equipment is built to auto sense what channels are most congested. Some access points will only perform this analysis when they start up, others will dynamically change their channel as needed. Between those two scenarios and manually specified channels, you can still run into situations where you experience heavy channel congestion. This is especially true in dense urban areas with lots of wireless networks in close proximity.

### Wireless Security

- **WEP (Wired Equivalent Privacy)**
An encryption technology that provides a very low level of privacy.

Using WEP protects your data a little but it should really only be seen as being as safe as sending unencrypted data over a wired connection. The WEP standard is a really weak encryption algorithm. It doesn't take very long for a bad actor to be able to break through this encryption and read your data. WEP only uses **40 bits** for its encryption keys and with the speed of modern computers, this can usually be cracked in just a few minutes. WEP was quickly replaced in most places with **WPA** or **Wi-Fi Protected Access**. WPA, by default, uses a **128-bit** key, making it a whole lot more difficult to crack than WEP. Today, the most commonly used encryption algorithm for wireless networks is **WPA2**, an update to the original WPA. WPA2 uses a **256-bit** key make it even harder to crack. Another common way to help secure wireless networks is through **MAC filtering**.
- **MAC filtering**
You configure your access points to only allow for connections from a specific set of MAC addresses belonging to devices you trust. This doesn't do anything more to help encrypt wireless traffic being sent through the air, but it does provide an additional barrier preventing unauthorized devices from connecting to the wireless network itself.

### Cellular Networking

Another super popular form of wireless networking is cellular networking, also called mobile networking. Cellular networks are now common all over the world. In some places, using a cellular network for Internet access is the most common way of connecting. At a high level, cellular networks have a lot in common with the 802.11 networks we've already talked about. Just like there are many different 802.11 specifications, there are lots of different cellular specifications. Just like Wi-Fi, cellular networking operates over radio waves, and there are specific frequency bands specifically reserved for cellular transmissions. One of the biggest differences is that these frequencies can travel over longer distances more easily, usually over many kilometers or miles.

Cellular networks are built around the concept of cells. Each cell is assigned a specific frequency band for use. Neighboring cells are set up to use bands that don't overlap, just like how we discussed the optimal setup for a W Lan with multiple access points. In fact, the cell towers that broadcast and receive cellular transmissions can be thought of like access points, just with a much larger range. Lots of devices today use cellular networks for communication. And not just phones, also tablets and some laptops also have cellular antennas. It's become more and more common for high-end automobiles to have built-in cellular access, too.

### Mobile Device Networks

Mobile devices use wireless networks to communicate with the Internet and with other devices. Depending on the device, it might use cellular networks, Wi-Fi, Bluetooth and or one of several Internet of Things or IoT network protocols.

It is also pretty common for a mobile device to have multiple network connections at the same time.
Mobile devices will use other **non-metered connections** like Wi-Fi, if they're available, so that you don't use up your limited data connection.

If you're troubleshooting an unreliable wireless network connection, keep in mind that wireless networking works by sending a radio signal between two **antennas**.

It might be printed on a circuit board, or it might have a wire or ribbon that runs through your device. The radio signal will get weaker the farther it has to travel, especially if it passes through or reflects off of things between the two antennas. Mobile devices can go with you to places where there is too much distance or interference for the wireless signal to be reliable. Even the way the mobile device is held or worn can impact the strength of the signal. So Wi-Fi and cellular data networks are used to connect your mobile devices to the internet. But there's one other type of wireless network to talk about. Mobile devices connect to their peripherals using short-range wireless networks. The most common short range wireless network is called Bluetooth.
When you connect a wireless peripheral to a mobile device, we call that pairing the devices. The two devices exchange information, sometimes including a PIN or password, so that they can remember each other. From then on, the devices will automatically connect to each other when they're both powered on and in range. Pairing devices like this can sometimes fail, and you might need to make your device forget the peripheral, so it can be paired again.