VoIP – How It’s Done

Introduction

Your business relies on your communication technology.  That’s why it is important for you to have some familiarity with the major concepts behind it.   You don’t have to be a tech-geek to grasp the broader concepts of voice over IP.   It can be intimidating,  but it’s not beyond your grasp.

Voice over IP uses the Internet to get your callers voice to your ear.  The Internet is where we have to start when learning about how VoIP works.  Everything hinges on something called the Internet Protocol, or IP.

In the paragraphs below, we touch on the most important concepts in ways that make sense without getting into technical jargon and all those confusing acrostics like H.323 and RFC 3611.

Digitizing, Packetizing, Transporting, Converting

Think of the Internet as the backbone of VoIP. Just as your spine connects all the parts of your body, the Internet is the connection between all parts of any VoIP system. Unlike your spine, however, the Internet is not a physical thing. While it is made up of physical parts in physical places, the Internet itself is abstract and constantly changing. The Internet is not the hardware, but how we use the hardware, copper wire, microwave links and fiber-optics.

The Packet: the most basic and critical component

The way we use the physical components of the Internet is by means of tiny fragments of information called packets. An IP packet is a small amount of digital data that is made up of a header, a payload or body, and trailer or footer. Each packet is a stand-alone unit, like a cell in your bloodstream. Each packet has an identity of sorts, so it can be recognized when it is mixed in with a group of other packets.

Creating Packets

When information is transmitted from person to another over the Internet, it is broken apart into packets, each containing some indentifying information and a portion of the information being transmitted.

With VoIP, like any digital audio, when you speak into the microphone, tiny electrical pulses are created. Those pulses are digitized by something called an Analog to Digital Converter or ADC.  The ADC creates a stream of electrical pulses representing the digits 1 and 0 (hence the term digital data, aka binary data).  Now that the voice is transformed into binary data, the computers between your caller and you can manipulate and manage it.

The digital stream is then broken up into tiny chunks of a certain size following the rules of the Internet Protocol, or IP. Each chunk of the stream become the body of a single IP packet.

The packet has a header, which identifies the packet. It contains information about who sent it, where it is going, a unique serial number and some other miscellaneous information.

The packet’s body contains a tiny portion of the digitized information being sent.

The packet’s trailer, or footer, contains a end-of-data marker and a CRC checksum, a special number which helps keep things together at the receiving end.

An Example for Scale

A stream of voice data contains thousands of packets. For example, I have an mp3 of Aerosmith’s “Dream On” on my hard drive. It plays 4 minutes 27 seconds. The file is 8,484,864 bytes in size. How many packets would it take to move it to your desk from mine?

There are 8 bits per byte, so that’s 67,878,912 bits of data contained in that song. If 128 bits of a 1024-bit packet are used by the header and footer, that leaves 896 bits for the body. So, transmitting this four minutes of music from one place to another would take 75,758 packets. (67,878,912 / 896 = 75,758)

By the time you add in the header and footer bits, the overall size of all the packets is 77,576,192 bits, or 9,697,024 bytes. All that math is intended to help you see how amazing it is that your data gets from one place to the other intact.

Getting Packets to the Receiver

Every packet leaves you with enough information to get where you wanted it to go all by itself. As soon as it leaves your home or office, each packet is dumped into a stream of millions of other packets traveling through the winding paths of copper, microwave and fiber-optics that make up the core network of the Internet. All sorts of equipment handles the packet as it moves along its journey. It gets routed, switched and sometimes replicated multiple times.

Each device that handles the packet either reads the header and decides what to do with it, or blindly treats it like any other packet and pushes it along its way. Routers and switches make decisions and the packet is sent in only one direction. Hubs and wifi repeaters do not make decisions. They broadcast the packets to everyone who hears them.

You can see how your packets can get into the hands of thousands of people, theoretically. And it does happen. Whether that is bad depends on what was in the body of the packet and who it is that has it. This brings up the topic of security, which is too much to discuss here.

Made It! Your Packet at its Destination.

Sooner or later, your packet finds its way to the intended receiver. It’s important to know that not every packet that left your computer took the same route to get to the receiver’s computer. Sometimes routers and switches have to send some packets a different way because of network loading. That means that sometimes the packets arrive out of order and have to be rearranged when they arrive.

The receiving device (computer) looks at the header, sees the serial number and slips it into the correct place in the line of packets it has received thus far. The body of the packet is stripped out and added to the end of the body from the previous packet. One by one, packet by packet, the original digital stream is re-created at the receiver’s computer.

If it is an audio stream, like your voice for instance, the digital stream is converted back into a stream of tiny electrical pulses, almost exactly like the ones created by your voice at the beginning. Those pulses then travel to a speaker of some kind where they are converted into sound, in this case the sound of your voice.

Why Do I Need to Know This?

When you understand what’s involved in getting your voice from your office to your friend on the other side of the ocean, you can better understand why you need business-class high speed internet service if you’re going to use VoIP in a large office.

You will also better understand why your residential VoIP service stutters a little when your children are playing online games and your significant other is streaming a movie while doing an online backup of her MP3 collection. It helps you better understand when things go wrong.

This brings us to bandwidth.

Packing Packets in a Pipe – Bandwidth

Believe it or not, a packet is a physical thing. It is made up of electrons moving in a stream that has physical restrictions. You can only pack so many packets in a given stream. The measure of how many packets can be in one stream at any given moment is known as bandwidth. Bandwidth is a critical concern when adding a new IP service to your Local Area Network (LAN). Every IP-based service you add consumes some portion of the available bandwidth you pay for each month.

Devices like IP security cameras and servers, video conference equipment, file servers and print servers all consume some part of your available bandwidth. In most settings, the majority of the impact is felt inside your office and many problems there can be remedied by reconfiguring your internal routing and switching hardware and cabling.

If you allow for online viewing of security camera systems, or if you run a web server or mail server that is accessible from outside your office, you are adding that outbound load to the web browsing and other inbound activities your office utilizes.

Your Internet service provider has limits on how much outbound and inbound traffic they allow based on your service level contract. This is often called upstream and downstream bandwidth. You can normally increase your available bandwidth by paying a higher monthly subscription.

One More Thing – Latency

Even light has a top speed. It takes a certain amount of time for a thing to travel from one place to another no matter what the thing is. The same is true with the packets in your VoIP communications. When packets disappear in-route the receiving end asks they be re-sent. This takes time. You may have seen what happens when packets are missing. It shows up as empty blocks on your digital television or even complete drop-outs.

The amount of time between the sending and receiving of a packet is called latency. A lot of work has gone into reducing latency in data systems including those serving the Internet. When the balance between bandwidth supply and demand tilts toward demand, latency tends to increase.

Latency causes awkward pauses to be felt in a communication situation. I speak to you. I hear my voice go out. If you’re in the same room, you hear me almost instantaneously. You think and you respond. I hear your response almost instantaneously. This feels natural.

When we’re using a tool like a satellite phone or the Internet to communicate, latency is introduced into the communications process. You speak. The words are digitized, packetized, transported, converted and then I hear your words. I think and respond. My words are digitized, packetized, transported, converted and you hear my response. There is a delay between your speech and my response. This may feel unnatural.

When the latency of the communication system is too great, it can be misunderstood as hesitation or unwillingness. This is where our perceptions are affected by our knowledge of the technology we are using. When you understand why latency happens, you are more likely to react well when it happens to you.

Latency should not normally be an issue for you if you are using VoIP. However, growing offices mean growing systems, and growing systems consume more and more resources. Latency often creeps up on you slowly until one day you suddenly notice it. Then you simply take measures to correct it. Contract for more bandwidth. Upgrade old equipment to faster equipment. Restructure your office intranet to give more internet resources to your VoIP system.

Hardware

There are currently three ways to use VoIP. You can use:

• your normal telephone equipment by adding an Analog Telephone Adapter (ATA).
• telephones that are designed for use with VoIP, also known as SIP (Session Initiated Protocol) phones.
• your laptop or desktop equipped with a microphone and software. This is called a softphone.

To be clear, all three options use SIP to enable and manage the VoIP connection.