The Basics of an Ethernet ...........................................................................
I can get deep into the technical details of the Ethernet here or I can acknowl-
edge that current versions of Windows and hardware are quite good at handling
all the hard work of configuration and setup. So let me not get too deep into
the details; I’ve included a little sidebar for those who really want to know a
bit about how a network works.200 Part IV: Failing to Communicate
Ethernet for those who have to know
The best way to think of Ethernet, or at least the
best one that I can wrap my arms around is this:
Think of a high-speed single-lane highway that
runs in a circle. Vehicles can get on or off the
road many places, but if they stay on the high-
way, they’ll eventually come back to where they
began and then start another circuit.
Now think of what you would have to do if you
were a driver trying to get onto the road; you
can’t just zoom onto the road without finding a
gap in the traffic large enough for your vehicle.
And then once you’re on the road you’ve got to
maintain speed while looking for the correct
exit, but you can only get off the road if there is
room on the off-ramp; if another vehicle is in
your way, you’ll have to go all the way around
the circle and try again. That’s an Ethernet, in a
very simplified description. Technically, a NIC
using that protocol employs Carrier Sense
Multiple Access With Collision Detection, a
fancy description for a scheme that monitors
the stream of traffic and looks for an opening.Information or instructions are broken up into
small pieces, called packets; the data is
enclosed within some extra coding that identi-
fies the sender and specifies the intended
recipient. When a NIC finds an opening in the
traffic, it sends out a packet; if two devices on
the network send a packet at the same moment,
there will be a collision.After a randomly
assigned wait, each computer is instructed to
resend the packet.I’ve described the Ethernet as a circle, a design
called a ring.In fact, the most common design for
an Ethernet is called a hub-and-spoke,with a
central hub or switch and each of the attached
computers located at the end of a spoke that
travels to and from its NIC. Another design is
called a bus,and it sets up with all of the com-
puters branching off a single main line.
The same concept is applied for wireless net-
works, which are set up very much like wired
systems — without the wires. A central wireless
hub receives all of the incoming information and
instructions from wireless computers and
retransmits packets to appropriate addresses.
Because the wireless design works almost the
same way as a wired Ethernet, an office or home
can create a network that intermixes devices
with and without wires.Today’s most common wired Ethernet systems in
homes and offices are based on cables and hard-
ware called 10Base-Tand capable of moving
data at as much as 10 megabits per second of
data, or 100Base-T which ups the speed to
100mbps. State-of-the-art is Gigabit Ethernet,
which can zoom along at as much as 1,000 Mbps.
Just like a single-lane highway, nothing can
travel on an Ethernet faster than the slowest
component on a particular link. If you have a
100Base-T hub and 10Base-T NICs, communi-
cation takes place at 10 Mbps.