Choosing a Network Connectivity Strategy

Now that we have taken a look at the commonly used LAN network architectures, let's concentrate on the different connectivity strategies that can be used to link your computers and other devices together. As you can see from our discussion of the various network architectures, the standard connectivity medium is some sort of cabling. The most popular cabling type is still copper wire. However, other technologies that make use of wireless technologies and wiring systems that are already in place (such as the telephone or electrical wiring system in a building) are also available and are becoming increasingly popular.

Let's take a look at some of the standards for network cables. Then we can look at some of the other connectivity strategies and how they are being used in different LAN settings.

Cabling Options

Network cabling consisting of copper wire has been the predominant network connectivity medium since the very beginning of the local area network. Copper-based cable transmits the data stream along the wire as an electrical signal. The discrete changes in the electrical signal on the cable distinguish the 1s and 0s in the bit stream. You already know that the network interface card (which we discussed in detail in the last chapter) has the job of taking the PC's information and getting it out onto the network medium in a format that is compatible with that particular medium (for example, electrical information is placed on copper wire, and light information is placed on fiber- optic cable).

Copper-based network cabling takes two major forms: coaxial cable and twisted-pair cable. Although the prices for fiber-optic cable (which does not experience interference because it uses light energy rather than electrical energy) are dropping and more fiber-optic LANs (particularly LAN backbones) are popping up all the time, copper wire is still the predominant wire type. Figure 4.3 provides a look at thinnet coaxial, thicknet coaxial, Category 5 twisted-pair cable, and fiber-optic cable, respectively.

Figure 4.3. Copper-based and fiber-optic cables are used as the physical media for LANs.

graphics/04fig03.gif

Let's take a closer look at copper-based possibilities. We can then take a look at fiber-optic cable.

Tip

graphics/tman.gif

When dealing with a physical medium such as copper wire (or fiber-optic cable for that matter), properties of the medium itself affect the distance that a signal can be sent and the integrity of that signal. Some physical phenomena you should know about include:

  • Attenuation. The degradation of the data signal over the run of the cable or medium (this is why different cable types have different maximum lengths).

  • Impedance. The resistance of a wire to data transmission. All cable types have impedance that is measured in Ohms. The greater the impedance, the more energy that is required to move the signal over the wire.

  • Interference. Signals from other nearby devices and noise from the wire itself (some of the energy on the wire is absorbed by the wire itself) can interfere with the actual transmission signal. In cases where two wires run closely together, they can interfere with each other's signal. This is called crosstalk .


Coaxial Cable

Two types of coaxial cable are used for networking computers: thicknet and thinnet. Thicknet is a heavy-gauge coaxial cable that is fairly inflexible and requires special equipment (over and above a simple network card) to connect the computer to the network backbone.

Installations of thicknet (RG-8 and RG-11 coaxial cable) are practically nonexistent, but some can still be found in certain settings, such as in manufacturing companies. This is because this thicker version of coaxial cable is well shielded (encased in a coating of material that looks somewhat like aluminum foil) and therefore doesn't suffer as much from interference as thinnet cable (a manufacturing environment with a lot of machinery in use can certainly produce potential interference problems when the cabling runs close to large electrical devices).

Thicknet networks are characterized by a cable backbone that is tied to servers and workstations on the network by vampire taps (the taps actually pierce the cable). The transceiver is actually attached to the tap (rather than a network interface card), and then the computer is connected to the transceiver/tap by a drop cable (which connects to the NIC on the computer).

Thinnet (RG-58 coaxial cable) was the cable of choice at one time because of its relative ease of installation and its low cost. Thinnet LANs employ a bus topology, where a T-connector is attached to each computer's network card. The computers are then chained together using appropriate lengths of cable. Thinnet installations require that each end of the network be terminated , and terminators are placed on the downside T-connector of the computers that reside on either end of the network.

Although it is valid for us to discuss coaxial cabling, in reality you are going to find that twisted-pair cable is the network medium of choice for most LANs. Let's take a look at twisted-pair wiring and how it works.

Twisted-Pair Wire

Twisted-pair wire comes in a number of different grades. Everyone is familiar with the twisted-pair wire used for the connection between your telephone and the phone wall outlet. Twisted pair is so named because the cable is actually made up of a number of single copper wires that are twisted together.

Why are the pairs of wire twisted? You already know that copper wire is susceptible to interference (as discussed in an earlier Tip sidebar). Two copper wires running parallel to each other would actually interfere with each other quite a bit. Believe it or not, by twisting the wires together, the interference (or crosstalk ) between the wires is minimized.

Twisted-pair wire comes in two major flavors: unshielded twisted pair (UTP) and shielded twisted pair (STP). The big difference between UTP and STP is that the STP wires are encased in a foil wrap that protects them from interference.

UTP is the most commonly used network wiring. It is inexpensive, flexible, and light, thus making it very easy to work with. UTP cable is terminated with an RJ-45 connector, which is kind of the big brother to the RJ-11 male connector that you find on a telephone cord.

UTP cable has been placed in different categories based on data-transmission capabilities. You will often hear UTP referred to as CAT 3 or CAT 5 . Table 4.1 provides a listing of these various UTP categories.

Table 4.1. Twisted Pair Categories

Category

Maximum Bandwidth Provided

Additional Information

1

None

Used in old telephone systems; this is not a data-grade cabling.

2

4Mbps

Not really considered a data-grade cable.

3

10Mbps

Considered the minimum cable requirement for data networks running Ethernet.

4

16Mbps

Equivalent to the Type 1 Token-Ring cabling without the shielding.

5

100Mbps

Has become the standard for new LAN installations and has completely overshadowed all the previous categories.

6

1,000Mbps

CAT 6 provides possibilities for broadband transmission. This may also end up being the copper cable of choice for Gigabit Ethernet networks.

Fiber-Optic Cable

Fiber-optic cable is a high-speed alternative to copper wire and is often employed as the backbone of larger corporate networks. However, the drop in the price of fiber-optic cable has started to make it a possibility for other LAN uses. Fiber-optic cable uses glass or plastic filaments to move data and provides greater bandwidth as well as longer cable runs (up to 2 kilometers, depending on the network architecture). With the need for network speed seemingly on the rise, fiber-optic installations are becoming commonplace.

Fiber-optic cable uses pulses of light as its data-transfer method. This means a light source is required, and lasers or light-emitting diodes are used. Fiber-optic cable is more expensive and more difficult to install than copper cable, but fiber-optic's ability to move data faster and farther makes it an excellent alternative to copper.

A very interesting aspect of fiber-optic cable is that because it uses light to transfer network data, it is not susceptible to tapping (that is, the stealing of data off the line itself). Copper cabling, on the other hand, can be tapped because it uses an electrical signal to send data. On networks where security is an issue, fiber-optic cable provides a more secure environment.

Note

graphics/nman.gif

Copper wire can be looped, bent, and put through all kinds of gyrations and still carry an electrical signal. Just think about how tangled your vacuum cleaner cord gets when you are vacuuming; however, the vacuum cleaner keeps running. Fiber-optic cable, on the other hand, has to remain relatively straight to carry data as a light pulse. You can't bend fiber-optic cable in a 45-degree angle and you can't have kinks in it. Fiber-optic network installations require cabling professionals who understand the fragility and maximum bend of fiber-optic cabling. This is why copper cable still reigns as king (or queen , for that matter); it is easy to install and fairly forgiving when it is installed poorly.


Phone and Electrical Wire Networks

One option for connecting PCs in a small network environment, such as a home peer-to-peer network or a small business network with a limited number of computers, is to use existing phone lines. Because the data communications on the network operate at a different frequency than the phone, phone communication is not disrupted by using the phone lines for a dual purpose (phone calls and the movement of data).

There is actually an organization called the Home Phoneline Networking Alliance (check out www.homepna.org for more information), which is attempting to create a standard for home networking products that use existing phone lines. Although many of the products available provide limited bandwidth, some supply 10Mbps throughput, which is as fast as the vanilla flavor of Ethernet. Phone line connectivity cards are available from a number of vendors . There is the Diamond PCI Phoneline card (it looks a lot like an Ethernet PCI NIC), called the Diamond Homefree. It can be used to connect PCs at 10Mbps by attaching them to any phone jack using a cable with RJ-11 ends (the same ends you find on any phone cord). Linksys also makes a phone line networking card. You will find that most of the phone-line products are compatible with the various flavors of Windows such as Windows 9x, 2000, and XP. Again, simple peer-to-peer networking is the aim of these products, as well as sharing a single Internet connection.

Other phone-line products take advantage of USB technology and use a small device that looks like a hub. The computers are attached to the device by a cable that runs from each computer's USB port. Your existing electrical lines also provide a source of wiring for connecting computers into a network. Again, a number of different companies make electrical or power-line networking products. This technology is considered to be another possibility for the home networking arena (as is phone-line technology) but could be used in some small business networking situations.

As with phone-line networking, using electrical lines as a connectivity medium allows computers to share files and printers and also share a single Internet connection. Power-line networks currently only offer bandwidth of around 350Kbps. This means that network communication is slower than some of the other alternatives (it is certainly not as fast as the phone-line alternatives).

Wireless Options

The options for creating wireless networks have seemingly exploded over the last couple years . Wireless network connections take advantage of radio signals, infrared light, or lasers. For longer distances, wireless communications can also take place through cellular telephone technology (as seen with any number of cellular devices, such as Web-ready cellular phones), through microwave transmission, or via satellite. Because the microwave and satellite technologies fall more into the realm of wide area networking, let's take a closer look at radio LAN technology and infrared technology. We will discuss cellular technology in Chapter 17, "Networking on the Run."

Radio Connectivity

Radio connectivity is used to extend the range of a LAN or make it easy for users with laptops to connect to the LAN without the use of cabling. Wireless radio LANs require that an access point be placed on the network for the computers that wish to connect to the LAN. For example, 3COM makes a wireless access point that attaches to an existing Ethernet network (Cisco also makes similar devices). It looks like a hub with an antenna on it. Computers outfitted with a wireless Ethernet card (these cards typically have a small antenna on them and otherwise operate like any other NIC using a transceiver) can connect to the access point. Figure 4.4 provides a diagram of a LAN that uses wireless technology for network access.

Figure 4.4. PCs outfitted with wireless network adapters connect to the LAN using a wireless access point.

graphics/04fig04.gif

The distance that a computer can be from the access point and still connect to the LAN varies from wireless product to wireless product. The 3COM specs for its wireless access point device claim a throughput of up to 11Mbps and a maximum range of 300 feet.

Wireless radio LANs are becoming increasingly popular in hospital settings (where nurses and doctors need to be mobile) and in business environments where a conference room must be quickly converted into a training room hosting multiple PCs connected to a LAN.

One issue related to wireless radio LANs is that it is fairly easy to intercept data. Although most radio LANs use one particular frequency, there are strategies for minimizing the possibility of data being intercepted (known as eavesdropping ).

Some wireless communications will actually break up the data and broadcast portions on different frequencies, thus making it harder to eavesdrop. Another method of diminishing the possibility of eavesdropping is known as frequency hopping . This is where the transmitter and the receiver switch (or hop ) frequencies periodically.

An important new set of specifications for wireless communication via radio signals is Bluetooth. A number of companies, including Microsoft, IBM, 3COM, and Nokia are creating Bluetooth devices such as keyboards, mice, personal digital assistants (PDAs), and cellular phones. Bluetooth is also being used to create networking devices.

Intel has certainly not been idle in relation to mobile computing and wireless communication. Intel has introduced the Centrino mobile computing technology, which integrates a wireless network connection into laptop computers that use the Intel Centrino processor.

Tip

graphics/tman.gif

Radio frequencies are regulated by the FCC in the United States. Therefore, to use most of the frequencies available in the radio wave spectrum, a license is required (for example, AM and FM radio stations must be licensed). Some frequencies, however, fall into an unregulated range. The FCC has reserved several frequency ranges for unregulated use (all measured in hertz): 902928MHz, 2.4GHz, and 5.725.83GHz. Cordless phones and remote control toys have sucked up most of the 900MHz range, so wireless LAN technologies typically operate at 2.4GHz and above.


Infrared Connectivity

Another possibility for wireless networking is infrared. Infrared systems use very high frequencies, just below visible light in the electromagnetic spectrum, to carry data. Like visible light, infrared signals can be blocked by opaque objects. Therefore, this technology is not suitable in all situations.

Because infrared technology is fairly limited in respect to the distance that two communicating devices can be apart, it is primarily used as a way for a laptop user to connect to a printer (both devices are outfitted with an infrared eye). Infrared is also becoming a popular method of syncing up personal data assistants, such as two Palm Pilots (we discuss handheld computing devices in Chapter 17, "Networking on the Run").

The Absolute Minimum

In this chapter we looked at the different network architectures that are used for local area networks. We also looked at the different connectivity mediums that can be used to connect networked devices such as copper cable, fiber optic cable, and wireless connection strategies.

  • Network architectures determine the physical layout of the network and the strategy used by computers to access the network medium.

  • Ethernet is a passive architecture and uses Carrier Sense Multiple Access with Collision Detection (CSMA/CD) as its network medium access strategy. It is defined by the IEEE 802.3 specifications. Faster versions of Ethernet, such as Fast Ethernet and Gigabit Ethernet, are also available.

  • IBM Token-Ring, which is defined by IEEE 802.5, uses possession of a token (which is passed around the logical network ring) as the network medium access method.

  • The two types of cable options for network media are copper cable and fiber-optic cable. Unshielded twisted-pair (UTP) cable is the cheapest and most often used cable type for new LANs.

  • Radio signals can be used to transfer data on wireless networks when an access point device is connected to the LAN and computers are outfitted with radio network cards.

  • Infrared is a wireless technology that is used to beam a signal between a laptop and a printer or to synchronize handheld personal data assistants.



Absolute Beginner's Guide to Networking
Absolute Beginners Guide to Networking (4th Edition)
ISBN: 0789729113
EAN: 2147483647
Year: 2002
Pages: 188
Authors: Joe Habraken

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net