
Ethernet Basics
Ethernet is a widely accepted, non-proprietary technology for networking computers and smart devices. Basically, Ethernet defines the communications that occur over a limited distance network. Ethernet was invented back in 1970 by some genius from Palo Alto by the name of Dr. Robert M. Metcalfe. Standards were developed in 1985, which are the basis for almost all Ethernet applications today. That standard is the IEEE802.3.
Basically, Ethernet is a technology for sending data in packets or frames. Essentially, Ethernet takes the data, whether it is a file, email, message, or data instructions, and puts it in a packet (also known as a frame.) Ethernet protocol establishes that all clients/stations/nodes (computers) on the network have equal access to the resources of the network. Basically, this means that any computer on the LAN can (and will) be sending data at the same time. Ethernet protocols specify how each computer will wait to send to listen for idle time on the medium. Once the NIC detects the channel is idle; it will send its packet.
Ethernet Packets consist of MAC (Medium Access Control) addresses, which is the hardware address pre-assigned by the manufacturer of the component, and IP header, an application address, the data being sent, and a trailer. To keep it simple, essentially the packet contains information about what is in the packet, where it came from, and where it has to go.
Since it is a shared network, a technology known as CSMA/CD (Carrier Sense Multiple Access with Collision Detect) is used to ensure only one client is sending at any one time. This protocol is used by the NIC cards and the Hub to determine if the cable (medium) is idle. When idle is detected, the data is sent. If there are simultaneous attempts by two clients on the LAN to send data, both back off, and retry at random intervals.
Another term used to describe Ethernet transmission is best effort packet sending, meaning the equipment does not guarantee success of packet transmissions as in more advanced technologies. On the other hand, for most LANs, this is not an issue.
Let’s get right down to it, how to wire, install, and configure a FAST Ethernet LAN:
What kind of LAN are you building? You're setting up a 100BaseT LAN, which is also known as FAST Ethernet, a network that provides 100mbs of bandwidth from client to client. (If you only wanted 10mbs, you'd be reading our previous discussion on Home Phone Line LANs.) You're not setting up a 100VG-Any LAN. 100VG-Any LAN is designed with upgrade paths to Token Ring and ATM technologies. For a home LAN, you don't need this. You'll have better compatibility with the more established technology, 100BaseT.
There are multiple types of Ethernet LANs, as listed below.
100BASE-TX –– Uses 2 pairs of Cat 5 balanced cable (balanced means both pairs the same length) 100BASE-FX –– Uses 2 multi mode fibers 100BASE-T4 –– Uses 4 pairs of Cat 3, 4, or 5 balanced cable. (designed to be used with Cat 3)
Types of cables listed below.
Category 3 This is a performance designation for twisted-pair cable and connecting hardware that can support frequency transmission up to 16 Mhz, and data rates of 10 Mbps. Category 3 has the capability to support low speed data applications, performing to the acceptable minimum for 100 ohm cabling systems; however it is now primarily used for telephone wiring.
Category 4 This category consists of cables and connectors specified up to 20 Mhz and data rates of 16 Mbps. Since the development of Category 5, however, Category 4 wiring systems are rarely used. Forget about this.
Category 5 This category consists of cables and connectors specified up to 100 Mhz and data rates of 100 Mbps, providing optimal performance for all data and phone systems. These systems are quickly becoming the standard because they provide a "safety net" to help ensure that current and future high-speed applications will run with peak accuracy, efficiency and throughput.
(Definitions above from www.levition.com)
All of the above cabling has a limitation of 100 meters. This shouldn’t be an issue. If you have a house big enough to need cable runs greater than 100 meters, you’re probably paying someone else to do this anyway.
You’re best bet, if you’re starting from scratch, is to wire it up with Category 5e (e=enhanced)cabling and jacks. Cat 3 cable is a stop gap measure, and cat 4 isn’t really used or supported. Regular Cat 5 is fine, but Cat 5e will give you an upgrade path to Gigabit Ethernet in the future. (As a gamer, you’ve got to like the promise of even more speed in the future!) Cat 5e will cost about 25-50% more in the cable and jacks, but worth it if you want the future capability.

The nice thing about Cat 5 wiring is everything is color coded, based on a standard. Thus, if you’’re not color blind, you should be able to pull this off. This is where you’ll need the punch down tool. You can buy a cheap one from a wiring supply store, or as suggested above, see if you can borrow one from work.
What are the differences between Category 3, 4, 5, etc. cable ratings?
Level 1: Basic telecommunications and power limited circuit cable.
Level 2: Up to to 1 MHz.
Category 3 (CAT3) is rated for transmission frequencies up to 16MHz. It is typically used for voice and data transmission rates up 10 Mbps (mega bits per second). Twisted-pair (not flat) phone wire.
CAT 4: frequencies up to 20 MHz... Voice and data transmission rates up to 16 Mbps.
CAT 5: frequencies up to 100 MHz... Voice and data transmission rates up to 100Mbps.
CAT 5e: frequencies up to 100 MHz. CAT 5e components have better transmission performance characteristics than CAT 5 and are most suitable for use in high-speed Gigabit Ethernets. Yes, 100 Mhz.
CAT 6 Supports transmissions up to 250 MHz. A standard for this category has not been finalized as of 5/21/01.
What is the difference between the ANSI/TIA/EIA T568A and T568B color code wiring standards?
T568A and T568B are the two color codes used for wiring eight-position RJ45 modular plugs. Both are allowed under the ANSI/TIA/EIA wiring standards. The only difference between the two color codes is that the orange and green pairs are interchanged. T568A wiring pattern is recognized as the preferred wiring pattern for this standard because it provides backward compatibility to both one pair and two pair USOC wiring schemes. The T568B standard matches the older ATA&T 258A color code and is/was(?) the most widely used wiring scheme. It is also permitted by the ANSI/TIA/EIA standard, but it provides only a single pair backward compatibility to the USOC wiring scheme. The U.S. Government requires the use of the preferred T568A standard for wiring done under federal contracts. The following diagrams look at the jacks from the front.
|
|
|
The wiring at the rear of the jack varies by manufacturer and may not be in the same sequence as the front. However, compliance with the color codes is maintained by routing the connections at the back to the proper sequence at the front of the jack. That is usually done by a small printed-circuit board in the jack assembly. CAT 5e jacks (right) may have a twist inside the jack to reduce crosstalk.
Again, please bear with me... Let's start with simple pin-out diagrams of the two types of UTP Ethernet cables and watch how committees can make a can of worms out of them.

Note that the TX (transmitter) pins are connected to corresponding RX (receiver) pins, plus to plus and minus to minus. And that you must use a cossover cable to connect units with identical interfaces. If you use a straight-through cable, one of the two units must, in effect, perform the cross-over function.
| RJ45 pin | name |
|---|---|
| 1 | transmit + |
| 2 | transmit - |
| 3 | receive + |
| 6 | receive - |
Cable Performance Testing In complex LANs with massive wiring, it is essential to have the cable tested to ensure it will support the applications. Here, we’re talking about a simple LAN, where this advanced testing is not necessary. If you’re interested in some of the testing, read the next paragraph, otherwise, go to the last section on connecting the whole works, so you get on to the important business of playing some on line games. Cat 3: 10Mbs, tested to 16Mhz Cat 5: 100Mbs, tested from 1Mhz to 100Mhz. The cable installer is responsible for ensuring documentation is complete and accurate of all cable runs, referred to as a wire map. In addition, attenuation, NEXT, and ACR are all tested using a cable meter that isn’t worth buying for a small job. For reference, companies such as Fluke, TTC (Tberd) and Hewlett Packard all make nice cable test sets. Attenuation is the measurement of signal loss from one end to the other, uses measured in decibels (db). The lower the signal loss, the better. NEXT measures Near End Cross Talk, which is a measurement of signal bleeding over from one cable to another. The measurement is actually the difference between the signal strength where it is supposed to be, and the residual on the adjacent cable, thus the higher the NEXT value, the better. Last, ACR is the attenuation to cross talk ratio, which isn’t actually a ratio, but the difference between the two. The higher the ACR rating, the better. This is accomplished by cabling with low loss, and very little cross talk. Connect and Configure You’re going to need to install NIC cards in each PC to be connected to the LAN. This is a piece of cake, especially with the Ethernet kits, such as the Dlink and Linksys systems. The cards are PCI cards, and it’s a plug and play operation to install them in your PC, no more complicated than installing a new sound card. Okey dokey, go ahead and connect the NICs to your jacks. You are probably using the RJ45 cables that came with the kit; usually you get two of these. These cables are likely straight through cables; ones that don’t flip transmit and receive. This is good, as the Hub does this for you. If you are not using a Hub, and connecting two PCs directly to each other, then you will have to use a "cross over" cable, one where the transmit and receive are flopped from one end to the other. Once you reboot, your PC should ask you for the drivers to install the new card. Using the disks or CDs provided should complete this task, but it is a good idea to check to ensure all necessary protocols have been installed. Those protocols are listed in the Control Panel, and can be viewed by double clicking the Network Icons. You want to see that Client for Microsoft Networks, Client for Netware Networks, the card driver you installed, IPX/SPX, NetBEUI, and TCP/IP are listed as installed. If any of the above are not, it is a simple task to install them. You can do so simply by clicking Add, and selecting Client for the first two, Adapter for you NIC, or Protocol for the last three. There’s a little more to it than listed, but its pretty straight forward. Your OS disk will have the necessary software to complete the installation of these protocols if needed. While you’re there, click on the Identification icon, and record the name of your computer. This is what that PC will be displayed as by the LAN. If your computer is called the ""Obliterator"", for example, that’s what it will read in the Computer Name section of this tab. If not named, go ahead and give this, and each computer on your LAN a unique name. Last, you’re going to have to set up an IP address for each PC on your LAN (if the software didn’t do it for you already.) Note, if one of your PCs is directly connected to the Internet, via DSL, cable, or ISDN, it may have a specific IP address, which you don’t want to change. For those PCs without specific IP addresses, double click again on the Network Icon in the control panel. Highlight TCP/IP, click properties, and select the IP address tab. Click the Specify an IP address tab, and then input the IP addresses for each PC. RFC 1918 defines a set of private IP addresses that are not to be routed on the Internet. These addresses include:
10.0.0.0/8 10.0.0.0_10.255.255.255 One Class A address
172.16.0.0/12 172.16.0.0_172.31.255.255 16 Class B addresses
192.168.0.0/16 192.168.0.0_192.168.255.255 256 Class C addresses
In addition, there are a number of reserved IP addresses that are never assigned to "public" networks, including:
0.0.0.0/32 Historical broadcast address
127.0.0.0/8 Loopback
169.254.0.0/16 Link_local Networks
192.0.2.0/24 TEST_NET
224.0.0.0/4 Class D Multicast address range
240.0.0.0/5 Class E Experimental address range
248.0.0.0/5 Unallocated
255.255.255.255/32 Broadcast
A good rule of thumb is to give the first one an IP address of xx.xx.xx.01, the next xx.xx.xx.02, etc.. For example, using 172.16.0.1, 172.16.0.2, ……, would work fine. For the subnet, use 255.255.255.0 for all the computers. You have now given each PC an individual address, thus allowing each to independently identified on your LAN. You may have to reboot each machine one more time to get them to be recognized by the LAN. Before trying this, double click on Network Neighborhood, and double click the Entire Network Globe. If this lists off all your PCs on the LAN, you’re pretty much good to go. This means the LAN is properly connected, and all PCs are recognized and able to transmit and receive. If not, try rebooting, and they should show up the next time you try this. All right, you are connected to the LAN, and Network Neighborhood shows all the PCs out there. You’ve pretty much finished the job (and probably that six pack I mentioned before.) Let’s test a few things. Try a simple ping test as mentioned in our last article. All you have to do is open a DOS window, and type ping, and the IP address of one of the PCs on your network. See below for a simple example.
C:\>ping 172.16.8.4
Pinging 172.16.8.4 with 32 bytes of data:
Reply from 172.16.8.4: bytes=32 time=25ms TTL=254
Reply from 172.16.8.4: bytes=32 time=11ms TTL=254
Reply from 172.16.8.4: bytes=32 time=21ms TTL=254
Reply from 172.16.8.4: bytes=32 time=17ms TTL=254
Ping statistics for 172.16.8.4:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 11ms, Maximum = 25ms, Average = 18ms
One more test to verify your new LAN is to set up some of the drives on your PCs as shared, allowing you to access them from other clients on your LAN. To set up your drives or printers as shared, again go to the Network icon in the control panel. This time, click on File and Print sharing, and click the boxes for drives and/or printers. Click OK, it will take you back to the Network window. Click on Add, Services, and File and Printer Sharing for Microsoft Networks. Go ahead and install this, it will ask you for your OS disk. Complete the software installation, and reboot. Once complete, all you have to do is option your hardware for sharing. You can do so from Windows Explorer, simply by right clicking on the drive, and selecting Sharing, and the option level you wish of access to the drives on your PC. Now that your drive is accessible to your network, you can test it by accessing it from another computer on your LAN, and do a couple of test copies of files from one PC to the other. Run some big files across your network to see how fast they go across. Keep in mind that although your LAN may be 100mbs, your hard drive may not be able to pump out the data that fast. On the other hand, if you can move a 100mb file in less than 10 seconds, it pretty safe to assume your LAN is running better than 10mbs. (Do the math.) For more intensive testing, many of the Home LAN kits have utilities for testing bandwidth, loop backs, and network settings. Again, the Dlink kit comes with several of these utilities, as an example. That, in a nutshell, is it. You now have connectivity to each of your PCs in equal fashion, meaning all PCs on your LAN share the same priority level. If you set it up, you can use one printer for the entire network, share files or hard drives, as well as share one high speed internet access. Best of all, though, is you can now have one raging game of head to head competition, without delays caused by network backups!
Glossary of Terms
ACR - Attenuation to Cross-Talk Ratio
ATM - Asynchronous Transfer Mode
CSMA/CD - Carrier Sense Multiple Access with Collision Detect
DHCP - Dynamic Host Control Protocol
EIA - Electronic Industries Alliance (or Association)
FDDI - Fiber Distributed Data Interface
IP - Internet Protocol
ISDN - Integrated Services Digital Network
ISP - Internet Service Provider
LAN - Local Area Network
MAC - Medium Access Control
MAN - Metro Area Network
MDI - Medium Dependent Interface
NIC - Network Interface Card
PCM - Pulse Code Modulation
PHY - Physical Layer Device
RJ-45 - 8 pin jack used to connect to the LAN
RSVP - Resource Reservation Protocol
RTP - Real Time Transfer Protocol
STP - Shielded Twisted Pair
TIA - Telecommunications Industry Association
TCP - Transfer Control Protocol
UDP - User Datagram Protocol
UTP - Unshielded Twisted Pair WAN - Wide Area Network
|
|
|