Ethernet Delivers the Internet

Most home-office or small-office situations bring the Internet into the local network through an Ethernet connection, leading to a basic router, interfaced with a DSL modem or a cable modem.

In order to understand the way Ethernet brings us the Internet, we have to take a little diversion into the concept of messages within messages.

Let me call your attention to a common, ordinary envelope. The outside of the envelope contains destination address and source address after the usual pattern that has been established for many, many years. In this case, the envelope is addressed to the Smith family, of 140 East Lilac, in Sunol, California. And, of course, the source address is placed in the upper left-hand corner according to well-established conventions.

It turns out that when a postman delivers this envelope to the Smith family according to this postal address, heís performing a function thatís pretty similar to the way Ethernet delivers frames using Ethernet addresses.

Our example will get a little more interesting as we open up the envelope and see whatís inside. Of course, opening the envelope and looking inside is something that the postman will generally not do. Heís only concerned with the address on the outside of the envelope.

So, opening up the envelope, we can find out whatís inside. And that reveals that thereís another package inside the first one; another envelope inside the envelope, representing a message inside a message. This time the addressing system is still present, but itís a little more intimate. It just says, ďTo Jessica, from Brian.Ē

Letís find out whatís inside by opening up this inner envelope.

We find there a CD with a name on it and a description of its contents on the outside. Opening up the CD jewel case will reveal that farther inside is a CD-ROM with additional information on what it contains. Of course, if we insert that CD into a CD player, weíll get some kind of an index of the songs on it.

As it turns out, we know something that the postman does not know, and that is that Jessica no longer lives here. So, letís close up the package and insert it back into the inner envelope with the personal address identifying Jessica and Brian, and weíll need to make a new outer envelope to address it to Jessicaís new, real address. The old envelope is simply discarded.

....There: a moment later, after addressing the new envelope to Jessicaís new address, weíll re-insert the inner package inside the new envelope, and send it on its way to Jessicaís new address; a message inside a message.

In the field of computer networking and Internet working, this concept of placing one package inside another is extended to placing one message inside another. And it is commonplace. Itís been going on for a long time, and this practice has become well-known under the name of ďMessage Encapsulation.Ē

Think back now to our first movie in our Ethernet series, the Ethernet Tutorial. Youíll remember that that tutorial revealed that a lot of the design for Ethernet is based on the behavior of telegraph operators. Remember that as telegraph operations matured, the format for messages became standardized. As telegraph operators sent a message, the first part of each message was a description of the destination telegraph office: the destination address to receive the telegram. After the destination address, the telegram included the source address. In essence, each telegram started with a preamble that said something like, ďTelegram to Salt Lake City, from DenverĒ. After those sections of a telegram identifying the receiving station, and the sending station, the main body of the text was transmitted.

Take a closer look at this telegram text.

Youíll see it contains a street address. This is a message within the telegram: a message within the message. And this additional addressing is necessary because telegraph addresses didnít go everywhere; every home did not have a telegraph receiver. A telegraph operator could address the city and get to an area, but ultimate delivery of the telegram contents to a personís home required an additional address. A boy on a bicycle would sometimes carry a hard-copy printout of a telegram to a personís house using the addressing information contained inside the telegram.

Well, Ethernet frames follow a pattern thatís very similar to telegrams. Inside the Ethernet frame, after the destination address, and the source Ethernet address, is the main message field. And that main message field inside the Ethernet frame can further be broken down into various sub-fields according to well-developed traditions and standards in the Ethernet community.

In fact, so many different people representing different organizations have had opinions and views and arguments about how to use the Ethernet message field that influential committees have been formed to create standards. Among the most important of these little sub-fields is one that tries to communicate a general description of the type of message and length of message transported inside this Ethernet frame. The exact formatting of the bits inside that little field have been a subject of considerable development and conflicting (some would say "confusing") evolution.

But itís not necessary that we understand those details. What is important is that we realize and trust that influential committees have agreed on the format by which information can be exchanged from one party to another, and that ultimately, our networking equipment can figure it out.

Probably the single most important message type that can be carried inside an Ethernet frame is an Internet protocol packet. This is the way Ethernet delivers the Internet.

I like to think of these committees and the protocols and standards that they create as if they are forms of the source that I use when I fill out my income tax reports every year. Thereís a place for every bit of information. And by following the form precisely, other people whoíve never met me can interpret details about my financial condition without very much confusion.

By following these well-established forms, or protocol conventions, the message field inside an Ethernet frame can carry many different kinds of messages. Our analogy with telegrams is valid here, too: Remember that the message inside a telegram often contained a postal address. In this same way, the IP packet, or Internet protocol packet, inside an Ethernet frame also contains an independent addressing system known as the IP address.

IP addresses are better-organized for delivery across the world in a world-wide collection of equipment and networks. It turns out that Ethernet addresses cannot be used effectively for world-wide distributions, because they are assigned more-or-less at random. Thereís no rhyme or reason in the local distribution of Ethernet addresses on your local network. The only requirement of Ethernet addresses is that they be unique. Imagine how difficult it would be for a world-wide postal service to address packages and envelopes if everybodyís address were completely random. Although an individual letter-carrier could probably memorize the few addresses on his local route, or carry a simple little map to help him sort out the local addresses, it would be completely unreasonable to expect world-wide postal delivery systems on a continental or inter-continental scale to be able to handle completely random address sorting. And thatís why Ethernet addresses cannot be used effectively outside your local area network. They just wonít scale to the world because they are distributed more-or-less at random.

IP addresses, on the other hand, are distributed in a more orderly fashion by an authority in charge of the organization of the Internet. And so the patterns in the numbers representing an Internet address help to locate the networks and sub-networks in the world-wide Internet. This notion of encapsulation is important to those of us with small networks, because when we connect our small networks to the Internet, we're actually merging two networks together. One Ethernet segment is actually owned and managed by our Internet service provider, and our own Ethernet is managed independently by our own policies and equipment.

And when an IP packet arrives from the Internet, we must remove it from the Ethernet frame managed by our ISP network, and repackage it inside a new Ethernet frame thatís understood by our local networking equipment for delivery to a computer inside our home or small business.

On the other hand, when we want to send a message out to someone on the world-wide Internet, our computers create an IP packet, which must be wrapped in an Ethernet frame for delivery to the Internet. The world-wide Internet is composed of millions of individual little networks, and every time an IP packet traverses any of them, it is encapsulated and then removed from framing such as that which we have illustrated here with Ethernet.

Ethernet Hubs and Ethernet Switches are not smart enough to do this. In order to look inside the Ethernet frame, find the internet protocol packet, interpret its structure, and find and handle the addressing, we use a smarter box known as a Router. Routers also allow us to share a single Internet address obtained from our Internet service provider, with multiple computers inside our home or small office network. You can learn a lot more about routers elsewhere on

This is the text from the movie entitled "Ethernet Delivers the Internet", published at (The text also includes a few small, static illustrations from the movie). You can learn more about the video version and purchase an inexpensive personal license to view it by clicking right here. You can learn more about our "Internet Series" of licensed videos by clicking here. Our "Internet Series" is part of a larger set of "Networking" videos, and you can learn about all of them by clicking here.