Like any good house, a network must be built on a solid foundation. In the OSI reference model, this foundation is called Layer 1 or the physical layer. The terms used in this chapter describe how network functions are linked to Layer 1 of the OSI reference model. The physical layer is the layer that defines the electrical, mechanical, procedural, and functional specifications for activating, maintaining, and deactivating the physical link between end systems.

In this chapter, you will learn about the network functions that occur at the physical layer of the OSI model. You will learn about different types of networking media that are used at the physical layer, including shielded twisted-pair cable, unshielded twisted-pair cable, coaxial cable, and fiber-optic cable. In addition, you will learn how network devices, cable specifications, network topologies, collisions and collision domains can help determine such things as how much data can travel across the network and how fast.

Shielded twisted-pair cable (STP) combines the techniques of shielding, cancellation, and twisting of wires . Each pair of wires is wrapped in metallic foil. The 4 pairs of wires are wrapped in an overall metallic braid or foil. It is usually 150 Ohm cable. As specified for use in Ethernet network installations, STP reduces electrical noise, both within the cable (pair to pair coupling, or crosstalk) and from outside the cable (electromagnetic interference -- EMI -- and radio frequency interference -- RFI). Shielded twisted-pair cable shares many of the advantages and disadvantages of unshielded twisted-pair cable (UTP). STP affords greater protection from all types of external interference, but is more expensive and difficult to install than UTP.

A new hybrid of UTP with traditional STP is Screened UTP (ScTP), also known as Foil Twisted Pair (FTP) . ScTP is essentially UTP wrapped in a metallic foil shield, or "screen". It is usually 100 or 120 Ohm cable.

The metallic shielding materials in STP and ScTP need to be grounded at both ends. If improperly grounded (or if there are any discontinuities in the entire length of the shielding material, for example due to poor termination or installation), STP and ScTP become susceptible to major noise problems, because they allow the shield to act like an antenna picking up unwanted signals. However, this effect works both ways. Not only does the foil (shield, screen) prevent incoming electromagnetic waves from causing noise on our data wires, but it minimizes the outgoing radiated electromagnetic waves, which could cause noise in other devices. STP and ScTP cable cannot be run as far as other networking media (coaxial cable, optical fiber) without the signal being repeated. More insulation and shielding combine to considerably increase the size, weight, and cost of the cable. And the shielding materials make terminations more difficult and susceptible to poor workmanship. However STP and ScTP still have their role, especially in Europe.

Unshielded twisted-pair cable (UTP) is a four-pair wire medium - composed of pairs of wires - used in a variety of networks. Each of the 8 individual copper wires in the UTP cable is covered by insulating material. In addition, each pair of wires are twisted around each other. This type of cable relies solely on the cancellation effect, produced by the twisted wire pairs, to limit signal degradation caused by EMI and RFI. To further reduce crosstalk between the pairs in UTP cable, the number of twists in the wire pairs varies. Like STP cable, UTP cable must follow precise specifications as to how many twists or braids are permitted per foot of cable.

When used as a networking medium, UTP cable has four pairs of either 22 or 24 gauge copper wire. UTP used as a networking medium has an impedance of 100 ohms. This differentiates it from other types of twisted-pair wiring such as that used for telephone wiring. Because UTP has an external diameter of approximately .43 cm, its small size can be advantageous during installation. Since UTP can be used with most of the major networking architectures, it continues to grow in popularity.

Unshielded twisted-pair cable has many advantages. It is easy to install and is less expensive than other types of networking media. In fact, UTP costs less per meter than any other type of LAN cabling, however its real advantage is its size. Since it has such a small external diameter, UTP does not fill up wiring ducts as rapidly as other types of cable. This can be an extremely important factor to consider, particularly when installing a network in an older building. Also, when UTP cable is installed using an RJ connector, potential sources of network noise are greatly reduced, and a good solid connection is practically guaranteed.

There are disadvantages in using twisted-pair cabling. UTP cable is more prone to electrical noise and interference than other types of networking media, and the distance between signal boosts is shorter for UTP than it is for coaxial and fiber optic cables.

While UTP was once considered slower at transmitting data than other types of cable. However, this is no longer true. In fact, today, UTP is considered the fastest copper-based media.

Coaxial cable consists of a hollow outer cylindrical conductor that surrounds a single inner wire made of two conducting elements. One of these elements - located in the center of the cable - is a copper conductor. Surrounding it is a layer of flexible insulation. Over this insulating material is a woven copper braid or metallic foil that acts as the second wire in the circuit, and as a shield for the inner conductor. This second layer, or shield, can help reduce the amount of outside interference. Covering this shield is the cable jacket.

For LANs, coaxial cable offers several advantages. It can be run, without as many boosts from repeaters, for longer distances between network nodes than either STP or UTP cable. Repeaters regenerate the signals in a network so that they can cover greater distances. Coaxial cable is less expensive than fiber-optic cable, and the technology is well known. It has been used for many years for all types of data communication. Can you think of another type of communication that utilizes coaxial cable?

When working with cable, it is important to consider its size. As the thickness, or diameter, of the cable increases, so does the difficulty in working with it. You must remember that cable must be pulled through existing conduits and troughs that are limited in size. Coaxial cable comes in a variety of sizes. The largest diameter was specified for use as Ethernet backbone cable because it had historically a greater transmission length and noise rejection characteristics. This type of coaxial cable is frequently referred to as thicknet. As its nickname suggests, this type of cable, because of its thickness, can be too rigid to install easily in some situations. The rule of thumb is: "the more difficult the network media is to install, the more expensive it is to install." Coaxial cable is more expensive to install than twisted-pair cable. Thicknet cable is almost never used anymore, except for special purpose installations.

In the past, coaxial cable with an outside diameter of only .35 cm (sometimes referred to as thinnet) was used in Ethernet networks. It was especially useful for cable installations that required the cable to make many twists and turns. Since it was easier to install, it was also cheaper to install. This led some people to refer to it as cheapernet. However, because the outer copper or metallic braid in coaxial cable comprises half the electrical circuit, special care must be taken to ensure that it is properly grounded. This is done by ensuring that there is a solid electrical connection at both ends of the cable. Frequently, installers fail to do this. As a result, poor shield connection is one of the biggest sources of connection problems in the installation of coaxial cable. Connection problems result in electrical noise that interferes with signal transmittal on the networking media. It is for this reason that, despite its small diameter, thinnet is no longer commonly used in Ethernet networks.

Fiber-optic cable is a networking medium capable of conducting modulated light transmissions. Compared to other networking media, it is more expensive; however, it is not susceptible to electromagnetic interference and is capable of higher data rates than any of the other types of networking media discussed here. Fiber-optic cable does not carry electrical impulses, as other forms of networking media that employ copper wire do. Instead, signals that represent bits are converted into beams of light. Even though light is an electromagnetic wave, light in fibers is not considered wireless because the electromagnetic waves are guided in the optical fiber. The term wireless is reserved for radiated, or unguided, electromagnetic waves.

Fiber-optic communication is rooted in a number of inventions made in the 19th century. It was not until the 1960s, when solid-state laser light sources and high-quality impurity-free glasses were introduced, that fiber-optic communication became practical. Its use on a widespread basis was pioneered by telephone companies, who saw its benefits for long-distance communication.

Fiber-optic cable used for networking consists of two fibers encased in separate sheaths. If viewed in cross section, you would see that each optical fiber is surrounded by layers of protective buffer material, usually a plastic such as Kevlar, and an outer jacket. The outer jacket provides protection for the entire cable. Usually made of plastic, it conforms to appropriate fire and building codes. The purpose of the Kevlar is to furnish additional cushioning and protection for the fragile hair-thin glass fibers. Wherever buried fiber-optic cables are required by codes, a stainless steel wire is sometimes included for added strength.

The light-guiding parts of an optical fiber are called the core and the cladding. The core is usually very pure glass with a high index of refraction. When the core glass is surrounded by a cladding layer of glass or plastic with a low index of refraction, light can be trapped in the fiber core. This process is called total internal reflection, and it allows the optical fiber to act like a light pipe, guiding light for tremendous distances, even around bends.

Wireless signals are electromagnetic waves, which can travel through the vacuum of outer space and through media such as air. Therefore, no physical medium is necessary for wireless signals, making them a very versatile way to build a network. Figure represents an electromagnetic wave.

Figure illustrates one of the most important charts in all of science and technology, the Electromagnetic Spectrum chart. You might be amazed that even tough all of the waves - power waves, radio waves, microwaves, Infrared light waves, visible light waves, ultraviolet light waves, x-rays, and gamma rays - look seemingly very different, they share some very important characteristics:

1. All of these waves have an energy pattern similar to that represented in Figure .
2. All of these waves travel at the speed of light, c = 299, 792, 458 meters per second, in vacuum. This speed might more accurately be called the speed of electromagnetic waves.
3. All of these waves obey the equation (frequency) x (wavelength) = c.
4. All of these waves will travel through vacuum, however, they have very different interactions with various materials.

The primary difference between the different electromagnetic waves is their frequency. Low frequency electromagnetic waves have a long wavelength (the distance from one peak to the next on the sine wave), while high frequency electromagnetic waves have a short wavelength.

The interactive calculator in Figure allows you to experiment. Try the interactive calculator by doing the following activities:

1. Enter a frequency and you will notice that the calculator displays the wavelength.
2. Enter a wavelength you will notice that the calculator displays the frequency.

In either case, the calculator display the electromagnetic wave associated with the calculation.

A common application of wireless data communication is for mobile use. Some examples of mobile use includes:

• people in cars or airplanes
• satellites
• remote space probes
• space shuttles and space stations
• anyone/anything/anywhere/anytime that requires network data
• communications, without having to rely on copper or optical fiber tethers

Another common application of wireless data communication is wireless LANs (WLANs), which are built in accordance with the IEEE 802.11 standards. WLANs typically use radio waves (for example, 902 MHz), microwaves (for example, 2.4 GHz), and Infrared waves (for example, 820 nanometers) for communication. Wireless technologies are a crucial part of the future of networking.

By the mid-1980s, growing pains from expansion in the field of networking were felt, especially by those companies that had instituted many different network technologies. It became increasingly difficult for networks that used different specifications and implementations to communicate with each other. An organization, called the International Organization for Standardization (ISO), researched various networks and created a network model, called the OSI Reference Model. (Note: Do not confuse the name of the model [OSI] with the name of the organization [ISO].) It was designed to help vendors create networks that would work compatibly and interoperably. By creating the OSI model, the ISO provided vendors with a set of standards.

Standards are sets of rules or procedures that are either widely used, or officially specified, and that serve as the gauge or model of excellence. The OSI model standards ensured compatibility and interoperability between the various types of network technologies that were produced by the many companies around the world. The early standards that were developed for networking media were largely proprietary. They were developed for use by various companies. Eventually, many more organizations and government bodies joined the movement to regulate and specify the types of cable that could be used for specific purposes or functions. Until recently, there has been a somewhat confusing mix of standards governing networking media. Standards have ranged from fire and building codes to detailed electrical specifications. Others have focused on tests to ensure safety and performance.

As you begin designing and building networks, make certain that you comply with all applicable fire codes, building codes, and safety standards. You should also follow any established performance standards in order to ensure optimal network operation and, because of the wide variety of options available today in networking media, to ensure compatibility and interoperability. Your work in this curriculum will focus on the standards for networking media that have been developed and issued by the following groups:

• IEEE - Institute of Electrical and Electronics Engineers
• UL - Underwriters Laboratories
• EIA - Electronic Industries Alliance
• TIA - Telecommunications Industry Association

The latter two organizations, jointly, issue a list of standards that you will frequently see listed as the TIA/EIA standards. In addition to these groups and organizations, local, state, county, and national government agencies issue specifications and requirements that can impact the type of cabling that can be used in a local area network.

The IEEE has outlined cabling requirements in its 802.3 and 802.5 specifications for Ethernet and Token Ring systems, and the standards for FDDI. Underwriters Laboratories issues cabling specifications that are primarily concerned with safety standards, however, they also rate twisted-pair networking media for performance. The Underwriters Laboratories established an identification program that lists markings for shielded and unshielded twisted-pair networking media in order to simplify the job of ensuring that materials used in LAN installations meet specifications

Of all of the organizations mentioned here, the TIA/EIA has had the greatest impact on networking media standards. Specifically, TIA/EIA-568-A and TIA/EIA-569-A, have been, and continue to be, the most widely used standards for technical performance of networking media.

The TIA/EIA standards specify the minimum  requirements for multi-product and multi-vendor environments. They allow for the planning and installation of LAN systems without dictating the use of specific equipment, thus giving LAN designers the freedom to create options for improvement and expansion.

The TIA/EIA standards address six elements of the LAN cabling process.  These are:

• horizontal cabling
• telecommunications closets
• backbone cabling
• equipment rooms
• work areas
• entrance facilities

This lesson will focus on TIA/EIA-568-A standards for horizontal cabling, which defines horizontal cabling as cabling that runs from a telecommunications outlet to a horizontal cross-connect. It includes the networking medium that runs along a horizontal pathway, the telecommunications outlet or connector, the mechanical terminations in the wiring closet, and the patch cords or jumpers in the wiring closet. In short, horizontal cabling includes the networking media that is used in the area that extends from the wiring closet to a workstation.

TIA/EIA-568-A contains specifications governing cable performance. It calls for running two cables, one for voice and one for data, to each outlet. Of the two cables, the one for voice must be four-pair UTP. The TIA/EIA-568-A standard specifies five categories in the specifications. These are category 1 (CAT 1), category 2 (CAT 2), category 3 (CAT 3), category 4 (CAT 4), and category 5  (CAT 5) cabling. Of these, only CAT 3, CAT 4, and CAT 5 are recognized for use in LANs. Of these three categories, CAT 5 is the one most frequently recommended and implemented in installations today.

The networking media that are recognized for these categories are the ones you have studied:

• shielded twisted-pair
• unshielded twisted-pair
• fiber-optic cable
• coaxial cable

For shielded twisted-pair cable, the TIA/EIA-568-A standard calls for two pair 150 ohm cable. For unshielded-twisted pair, the standard calls for four pair 100 ohm cable. For fiber-optic, the standard calls for two fibers of 62.5/125  multi-mode cable. Although 50 ohm coaxial cable is a recognized type of networking media in TIA/EIA-568-A, it is not recommended for new installations. Moreover, this type of coaxial cable is expected to be removed from the list of recognized networking media the next time this standard is revised.

For the horizontal cabling component, TIA/EIA-568A requires a minimum of two telecommunications outlets or connectors at each work area. This telecommunications outlet/connector is supported by two cables. The first is a four-pair 100 ohm CAT 3 or higher UTP cable along with its appropriate connector. The second can be any one of the following:

• four-pair 100 ohm unshielded twisted-pair cable and its appropriate connector
• 150 ohm shielded twisted-pair cable and its appropriate connector 
• coaxial cable and its appropriate connector
• two-fiber 62.5/125 µ optical fiber cable and its appropriate connector

According to TIA/EIA-568-A, the maximum distance for cable runs in horizontal cabling is 90 meters (m). This is true for all types of CAT 5 UTP recognized networking media. The standard also specifies that patch cords or cross-connect jumpers located at the horizontal cross-connect cannot exceed 6 m in length. TIA/EIA-568-A also allows 3 m for patch cords that are used to connect equipment at the work area. The total length of the patch cords and cross-connect jumpers used in the horizontal cabling cannot exceed 10 m. A final specification for horizontal cabling contained in TIA/EIA-568-A requires that all grounding and bonding must conform to TIA/EIA-607 as well as to any other applicable codes.

The latest industry standards being developed are for Cat 5e, Cat 6, and Cat 7 cabling, all of which offer improvements over Cat 5

Cables must eventually be terminated in order to provide connectivity. This process involves much transition and innovation as far as computer networking is concerned. This presents a tremendous challenge for students, who must learn a wide variety of networking medium standards, properties, and terminations.

Explore the following Web sites to examine the variety of cable and termination products that are available.

In this curriculum you will be introduced to three LAN technologies: Ethernet, Token Ring, and FDDI. All three have a wide variety of Layer 1 components and devices. The focus of this chapter will be Ethernet 10BASE-T technologies.

When it was developed, Ethernet was designed to fill the middle ground between long distance, low-speed networks, and specialized computer room networks that carried data at high speeds for very limited distances. Ethernet is well-suited to applications in which a local communication medium must carry sporadic, occasionally heavy, traffic at high-peak data rates.

The Ethernet 10BASE-T technologies carry Ethernet frames on inexpensive twisted-pair wiring. You will study four components and three devices that are related to these technologies. The first four components are passive, meaning they require no energy to operate. They are:

• patch panels
• plugs
• cabling
• jacks

The last three are active. They require energy to do their jobs. They are:

• transceivers
• repeaters
• hubs

For more information on other Ethernet, Token Ring, and FDDI components and devices, go to the Web site

The standard 10BASE-T termination (end point, 0 plug, connector) is the registered jack-45 connector (RJ-45).  It reduces noise, reflection, and mechanical stability problems, and resembles a phone plug, except that it has eight conductors instead of four. It is considered a passive networking component because it only serves as a conducting path between the four pairs of Category 5 twisted cable and the prongs of the RJ-45 jack. It is considered a Layer 1 component, rather than a device, because it serves only as a conducting path for bits.

The standard 10BASE-T cable is CAT 5 twisted-pair cable, which is composed of four twisted pairs that reduce noise problems. CAT 5 is thin, inexpensive, and easy to install. The function of CAT 5 cable is to carry the bits, therefore, it is a Layer 1 component.

RJ-45 plugs fit into RJ-45 jacks or receptacles. The RJ-45 jack has eight conductors , which snap together with the RJ-45 plug. On the other side of the RJ-45 jack is a punch down block where wires are separated out and forced into slots with a fork-like tool called a punch-down tool. This provides a copper-conducting path for the bits. The RJ-45 jack is a Layer 1 component.

Patch panels are convenient groupings of RJ-45 jacks. They come in 12, 24, and 48 ports, and are typically rack-mounted. The front sides are RJ-45 jacks ; the back sides are punch-down blocks that provide connectivity or conducting paths. They are classified as Layer 1 devices.

A transceiver is a combination of transmitter and receiver. In networking applications, this means that they convert one form of signal to another form. For example, many networking devices come with an auxiliary unit interface and a transceiver that allows a 10BASE2, 10BASE5, 10BASE-T, or 10\100 BASE-FX to be connected to the port. A common application is the conversion of AUI ports to RJ-45 ports. They are Layer 1 devices. They transmit from one pin configuration and/or media to another. Transceivers are often built into NICs, which are typically considered Layer 2 devices. Transceivers on NICs are called signaling components, which means they encode signals onto the physical medium.

Repeaters regenerate, and retime signals, which then enables cables to extend farther to reach longer distances. They only deal with packets at the bit level, therefore they are Layer 1 devices.

Repeaters are internetworking devices that exist at the physical layer (Layer 1) of the OSI model. They can increase the number of nodes that can be connected to a network, and thus, the distance over which the network can extend. Repeaters re-shape, regenerate, and retime signals before sending them on along the network.

The disadvantage of using repeaters is that they cannot filter network traffic. Data (bits) that arrive at one port of a repeater are sent out on all other ports. The data gets passed along to all other LAN segments of a network regardless of whether or not it needs to go there.

Multiport repeaters combine connectivity with the amplifying and re-timing properties of repeaters. It is typical to see 4, 8, 12, and up to 24, ports on multiport repeaters. This allows many devices to be cheaply and easily interconnected. Multiport repeaters are often called hubs, instead of repeaters, when referring to the devices that serve as the center of a star topology network. Hubs are very common internetworking devices. Since the typical unmanaged hub only requires power and plugged-in RJ-45 jacks, they are great for setting up a network quickly. Like the repeaters on which they are based, they only deal with bits, and are Layer 1 devices.

All of these devices - passive and active - create or act on the bits. They recognize no information patterns in the bits, no addresses, and no data. Their function is simply to move bits around. Layer 1 is fundamental to troubleshooting networks, and should never be underestimated. Many network problems are traceable to bad RJ-45 terminations, jacks, punch-downs, repeaters, hubs, or transceivers.

Collisions and Collision Domains in Shared Layer Environments

Some networks are directly-connected; all hosts share Layer 1. Examples are:

·         shared media environment - occurs when multiple hosts have access to the same medium. For example, if several PCs are attached to the same physical wire, optical fiber, or share the same airspace, they all share the same media environment. Occasionally you may hear someone say "all the computers are on the same wire" . It means that they all share the same media - even though the "wire" might be CAT 5 UTP, which has four pairs of wire.

·         extended shared media environment - is a special type of shared media environment in which networking devices can extend the environment so that it can accommodate multiple-access, or more users. There are, however, negative aspects to this as well as positive aspects.

·         point-to-point network environment - is most widely used in dial-up network connections, and is the one with which you are most likely familiar. It is a shared networking environment in which one device is connected to only one other device via a link, such as you connecting to internet service provider by phone line.

Some networks are indirectly-connected, meaning that some higher layer networking devices and/or some geographical distance is between the two communicating hosts. There are two types:

·         circuit-switched - an indirectly-connected network in which actual electrical circuits are maintained for the duration of the communication. The current telephone system is still, in part, circuit-switched, although the telephone systems in many countries is now concentrating less on circuit-switched technologies.

·         packet-switched - rather than dedicating a link as an exclusive circuit connection between two communicating hosts, the source sends messages in packets. Each packet contains enough information for it to be routed to the proper destination host. The advantage is that many hosts can share the same link; the disadvantage is that conflicts can occur.

Collisions and Collision Domains in Shared Layer Environments

Some networks are directly-connected; all hosts share Layer 1. Examples are:

·         shared media environment - occurs when multiple hosts have access to the same medium. For example, if several PCs are attached to the same physical wire, optical fiber, or share the same airspace, they all share the same media environment. Occasionally you may hear someone say "all the computers are on the same wire" . It means that they all share the same media - even though the "wire" might be CAT 5 UTP, which has four pairs of wire.

·         extended shared media environment - is a special type of shared media environment in which networking devices can extend the environment so that it can accommodate multiple-access, or more users. There are, however, negative aspects to this as well as positive aspects.

·         point-to-point network environment - is most widely used in dial-up network connections, and is the one with which you are most likely familiar. It is a shared networking environment in which one device is connected to only one other device via a link, such as you connecting to internet service provider by phone line.

Some networks are indirectly-connected, meaning that some higher layer networking devices and/or some geographical distance is between the two communicating hosts. There are two types:

·         circuit-switched - an indirectly-connected network in which actual electrical circuits are maintained for the duration of the communication. The current telephone system is still, in part, circuit-switched, although the telephone systems in many countries is now concentrating less on circuit-switched technologies.

·         packet-switched - rather than dedicating a link as an exclusive circuit connection between two communicating hosts, the source sends messages in packets. Each packet contains enough information for it to be routed to the proper destination host. The advantage is that many hosts can share the same link; the disadvantage is that conflicts can occur.

Collisions and Collision Domains in Shared Layer Environments

A situation that can occur, when two bits propagate at the same time on the same network, is a collision. A small, slow network could work out a system that allowed only two computers to send messages, each agreeing to take turns. That would mean that they could both send messages, but there would be only one bit on the system. The problem is that many computers are connected to large networks, each one wanting to communicate billions of bits every second. It's also important to remember that the "bits" are actually packets containing many bits.

Serious problems can occur as a result of too much traffic on a network. If there is only one cable that interconnects all of the devices on a network, the possibility of conflicts with more than one user sending data at the same time is very high. The same is true if segments of a network are only connected by non-filtering devices, such as repeaters. Ethernet allows only one data packet to access the cable at any one time. If more than one node attempts to transmit at the same time, a collision occurs, and the data from each device suffers damage .

The area within the network, where the data packets originated and collided, is called a collision domain, and includes all shared media environments. One wire may be connected to another wire through patch cables, transceivers, patch panels, repeaters, and even hubs. All of these Layer 1 interconnections are part of the collision domain.

Collisions and Collision Domains in Shared Layer Environments

When a collision occurs, the data packets that are involved are destroyed, bit by bit. In order to avoid this problem, the network should have in place a system that can manage the competition for the medium (contention). For example, a digital system can only recognize two voltage, light, or electromagnetic wave states. Therefore, in a collision, the signals interfere, or collide, with each other. Just as two cars cannot occupy the same space, on the same road, at the same time, neither can two signals occupy the same medium, at the same time.

Collisions and Collision Domains in Shared Layer Environments

Normally, people think that collisions are bad because they decrease network performance. However, a certain amount of collisions are a natural function of a shared media environment (i.e. collision domain). This is because large numbers of computers are all trying to communicate with each other at the same time, by using the same wire. 

The history of how Ethernet handles collisions and collision domain dates back to research at the University of Hawaii. In its attempts to develop a wireless communication system for the Islands of Hawaii, university researchers developed a protocol called Aloha. This protocol was instrumental in the development of Ethernet.

Collisions and Collision Domains in Shared Layer Environments

As a networking professional, one important skill is the ability to recognize collision domains. If you connect several computers to a single medium that has no other networking devices attached, you have a shared-access situation, and you have a collision domain. Depending on the particular technology used, this situation limits the number of computers that can use that part of the medium, also called a segment.

Collisions and Collision Domains in Shared Layer Environments

Repeaters regenerate and retime bits, but they cannot filter the flow of traffic that pass through them. Data (bits) that arrive at one port of a repeater are sent out on all other ports. Using a repeater extends the collision domain, therefore, the network on both sides of the repeater is one larger collision domain.

Collisions and Collision Domains in Shared Layer Environments

You have already learned that another name for a hub is a multiport repeater. Any signal that comes in one port of the hub is regenerated, retimed, and sent out every other port. Therefore, hubs, which are useful for connecting large numbers of computers, extend collision domains. The final result is diminished network performance if all the computers on that network are demanding large bandwidths, simultaneously.

Collisions and Collision Domains in Shared Layer Environments

Both repeaters and hubs are Layer 1 devices, therefore they perform no filtering of network traffic. Extending a run of cable with a repeater, and ending that run with a hub, results in a larger collision domain.

Collisions and Collision Domains in Shared Layer Environments

The four repeater rule in Ethernet states, that no more than four repeaters or repeating hubs can be between any two computers on the network. To assure that a repeated 10BASE-T network will function properly, the following condition must be true: (repeater delays + cable delays + NIC delays) x 2 < maximum round-trip delay. Repeater delays for 10BASE-T are usually less than 2 microseconds per repeater; cable delays are near 0.55 microseconds per 100 m trip; NIC delays are about 1 microsecond per NIC; and the maximum round-trip delay (the 10BASE-T bit time of 0.1 microseconds times the minimum frame size of 512 bits) is 51.2 microseconds. For a 500 m length of UTP connected by 4 repeaters (hubs) and 2 NIC delays the total delay would be well below the maximum round-trip delay. Repeater latency, propagation delay, and NIC latency all contribute to the 4-repeater rule. Exceeding the four repeater rule can lead to violating the maximum delay limit.

When this delay limit is exceeded, the number of late collisions dramatically increase. A late collision, is when a collision happens after the first 64 bytes of the frame are transmitted. The chipsets in NICs are not required to retransmit automatically when a late collision occurs. These late collision frames add delay referred to as consumption delay. As consumption delay and latency increase, network performance decreases. This Ethernet rule of thumb is also known as the 5-4-3-2-1 rule. Five sections of the network, four repeaters or hubs, three sections of the network are "mixing" sections (with hosts), two sections are link sections (for link purposes), and one large collision domain.

Collisions and Collision Domains in Shared Layer Environments

Although repeaters and hubs are useful, inexpensive networking devices, they extend collision domains. If the collision domain becomes too large, this can cause too many collisions and result in poor network performance.   The size of collision domains can be reduced by using intelligent networking devices that break up the domains. Examples of this type of networking device are bridges, switches, and routers. This process is called segmentation.

A bridge can eliminate unnecessary traffic on a busy network by dividing a network into segments and filtering traffic based on the station address. Traffic between devices on the same segment does not cross the bridge, and does not affect other segments. This works well as long as the traffic between segments is not too heavy. Otherwise, the bridge can become a bottleneck, and actually slow down communication.

The word topology can be thought of as "the study of location." Topology is a subject of study in mathematics, where maps of nodes (dots) and links (lines) often contain patterns. In this chapter, you will examine the various topologies used in networking from a mathematical perspective. Then, you will learn how a physical topology describes the plan for wiring the physical devices. Finally, you will use a logical topology to learn how information flows through a network to determine where collisions may occur.

A network may have one type of physical topology , and a completely different type of logical topology. For example, Ethernet 10BASE-T uses an extended-star physical topology, but acts as though it uses a logical bus topology. Token Ring uses a physical star, and a logical ring. FDDI uses a physical and a logical ring.

Mathematical Perspective
The bus topology has all of its nodes connected directly to one link, and has no other connections between nodes

Physical Perspective
Each host is wired to a common wire. In this topology, the key devices are those that allow the host to join or tap into the single shared medium. One advantage of this topology is that all hosts are connected to each other, and thus, can communicate directly. One disadvantage of this topology is that a break in the cable disconnects hosts from each other.

Logical Perspective
A bus topology enables every networking device to see all signals from all other devices. This can be an advantage if you want all information to go to every device.

Mathematical Perspective
A ring topology is a single closed ring consisting of nodes and links, with each node connected to only two adjacent nodes.

Physical Perspective 
The topology shows all devices wired directly to each other in what is called a daisy-chain. This is similar to the manner in which a mouse on an Apple PC plugs into the keyboard and then into the PC.

Logical Perspective
In order for information to flow, each station must pass the information to its adjacent station.

Mathematical Perspective
A dual ring topology consists of two concentric rings, each of which is linked only to its adjacent ring neighbor. The two rings are not connected.

Physical Perspective
A dual ring topology is the same as a ring topology, except that there is a second, redundant ring, that connects the same devices. In other words, in order to provide reliability and flexibility in the network, each networking device is part of two independent ring topologies.

Logical Perspective
A dual ring topology acts like two independent rings, of which, only one at a time is used.

Mathematical Perspective
A star topology has a central node with all links to other nodes radiating from it and allows no other links.

Physical Perspective
A star topology has a central node with all links radiating from it. Its primary advantage is that it allows all other nodes to communicate with each other, conveniently. Its primary disadvantage is that if the central node fails, the whole network becomes disconnected. Depending on the type of networking device used at the center of the star network, collisions can be a problem.

Logical Perspective
The flow of all information would go through one device. This might be desirable for security or restricted access reasons, but it would be very susceptible to any problems in the star's central node.

Mathematical Perspective
An extended star topology repeats a star topology, except that each node that links to the center node is, also, the center of another star.

Physical Perspective
An extended star topology has a core star topology, with each of the end nodes of the core topology acting as the center of its own star topology. The advantage of this is that it keeps wiring runs shorter, and limits the number of devices that need to interconnect to any one central node.

Logical Perspective
An extended star topology is very hierarchical, and information is encouraged to stay local. This is how the phone system is currently structured.

Mathematical Perspective
The tree topology is similar to the extended star topology, the primary difference is that it does not use one central node. Instead, it uses a trunk node from which it branches to other nodes. There are two types of tree topologies: the binary tree (each node splits into two links); and the backbone tree (a backbone trunk has branch nodes with links hanging from it).

Physical Perspective
The trunk is a wire that has several layers of branches.

Logical Perspective
The flow of  information is hierarchical.

Mathematical Perspective
In the irregular network topology there is no obvious pattern to the links and nodes.

Physical Perspective
The wiring is inconsistent; the nodes have varying numbers of wires leading from them. This is how networks that are in the early stages of construction, or poorly planned, are often wired.

Logical Perspective
There is no obvious pattern to the links and nodes.

Mathematical Perspective
In a complete, or mesh topology, every node is linked directly to every other node.

Physical Perspective
This wiring has very distinct advantages and disadvantages. One advantage is every node is physically connected to every other node (creating a redundant connection). Should any link fail to function, information can flow through any number of other links to reach its destination. Another advantage of this topology is that it allows information to flow along many paths on its way back through the network. The primary physical disadvantage is that for anything more than a small number of nodes, the amount of media for the links, and the amount of connections to the links becomes overwhelming.

Logical  Perspective
The behavior of a complete, or mesh topology depends greatly on the devices used.

Mathematical Perspective
The cellular topology consists of circular or hexagonal areas, each of which has an individual node at its center.

Physical Perspective
The cellular topology is a geographic area that is divided into regions (cells) for the purposes of wireless technology – a technology that becomes increasingly more important each day. There are no physical links in a cellular topology, only electromagnetic waves. Sometimes the receiving nodes move (e.g. car cell phone), and sometimes the sending nodes move (e.g. satellite communication links).

The obvious advantage of a cellular (wireless) topology is that there are no tangible media other than the earth's atmosphere or the vacuum of off-planet space (and satellites). The disadvantages are that signals are present everywhere in a cell and, thus, are susceptible to disruptions (man-made and environmental) and to security violations (i.e. electronic monitoring and theft of service).

Logical Perspective
Cellular technologies communicate with each other directly (though distance limitations and interference sometimes make it extremely difficult), or communicate only with their adjacent cells, which is  extremely inefficient. As a rule, cellular-based topologies are integrated with other topologies, whether they use the atmosphere or satellites.

In this chapter, you learned that the function of the physical layer is to transmit data. In addition you learned that the following types of networking media may be used to connect computers:

• Coaxial cable consists of a hollow outer cylindrical conductor that surrounds a single inner wire conductor.
• UTP cable is a four-pair wire medium used in a variety of networks.
• STP cable combines the techniques of shielding, cancellation, and twisting of wires.
• Fiber-optic cable is a networking medium capable of conducting modulated light transmissions.

This chapter discussed various criteria, such as rate of data transfer and expense, that help determine which types of networking media should be used. You learned that TIA/EIA-568-A and TIA/EIA-569 are the most widely used standards for technical performance of networking media.

Additionally, this chapter described how bits propagating at the same time on the same network result in a collision.  Lastly, you learned that a network may have one type of physical topology, and a completely different type of logical topology.  In the next chapter, you will learn about LAN media and the IEEE model and how the data link layer provides reliable transit of data across a physical link by using Media Access Control (MAC) addresses.