Lesson 1: Connectivity

Devices
Ing. José L. Simancas García, Mg.
Modem Technology
A modem is a device that makes it possible for computers to communicate over a telephone line. When computers are too far
apart to be joined by a standard computer cable, a modem can enable communication between them.
In a network environment, modems serve as a means of communication between networks and as a way to connect to the world
beyond the local network.
Computers cannot simply be connected to
each other over a telephone line, because
computers communicate by sending digital
electronic pulses (electronic signals), and a
telephone line can send only analog waves
(sound). Figure 7.1 shows the difference
between digital computer communication
and analog telephone communication.

A digital signal has a binary form. The signal

can have a value of either 0 or 1. An analog
signal can be pictured as a smooth curve that
can represent an infinite range of values.
As shown in Figure 7.2, the modem at the sending end converts the computer's digital signals into analog waves and transmits
the analog waves onto the telephone line. A modem at the receiving end converts the incoming analog signals back into digital
signals for the receiving computer.
In other words, a sending modem MOdulates digital signals into analog signals, and a receiving modem DEModulates analog
signals back into digital signals.
Modem Hardware
Modems are known as data communications equipment (DCE) and share the following characteristics:
•A serial (RS-232) communications interface
•An RJ-11 telephone-line interface (a four-wire telephone plug)

Modems are available in both internal and external

models. An internal modem, as shown in Figure 7.3,
is installed in a computer's expansion slot like any
other circuit board.

An external modem, as shown in Figure

7.4, is a small box that is connected to
the computer by a serial (RS-232) cable
running from the computer's serial port
to the modem's computer cable
connection. The modem uses a cable
with an RJ-11C connector to connect to
the wall.
Modem Standards
Standards are necessary so that modems from one manufacturer can communicate with modems from another manufacturer.
This section explains some of the common industry standards for modems.
Hayes-Compatible
In the early 1980s, a company called Hayes Microcomputer Products developed a modem called the Hayes Smartmodem. The
Smartmodem became the standard against which other modems were measured, and generated the phrase "Hayes-compatible,"
just as IBM's personal computer generated the term "IBM-compatible." Because most vendors conformed to the Hayes
standards, nearly all LAN modems could communicate with each other.
The early Hayes-compatible modems sent and received data at 300 bits per second (bps). Modem manufacturers currently offer
modems with speeds of 56,600 bps or more.

International Standards
Since the late 1980s, the International Telecommunications Union (ITU) has developed standards for modems. These
specifications, known as the V series, include a number that indicates the standard. As a reference point, the V.22bis modem at
2400 bps would take 18 seconds to send a 1000-word letter. The V.34 modem at 9600 bps would take only four seconds to send
the same letter, and the V.42bis compression standard in a 14,400 bps modem can send the same letter in only three seconds.

The chart in Table 7.1 presents the compression standards and their parameters since 1984. The compression standard and the
bps are not necessarily related. The standard could be used with any speed of modem.
Modem Performance
Initially, a modem's speed was measured in either bps or something called the "baud rate," and most people mistakenly
assumed the two were identical.
"Baud" refers to the speed at which the sound wave that carries a bit of data over the telephone lines oscillates. The term
derives from the name of French telegrapher and engineer Jean-Maurice-Emile Baudot. In the early 1980s, the baud rate did
equal the transmission speed of modems. At that time, 300 baud equaled 300 bits per second.
Eventually, communications engineers learned to compress and encode data so that each modulation of sound could carry more
than one bit of data. This development means that the rate of bps can be greater than the baud rate. For example, a modem
that modulates at 28,800 baud can actually send at 115,200 bps. Therefore, the current parameter to look for in modem speed is
bps.
Several of the newer modems feature industry standards, such as V.42bis/MNP5 data compression, and have transmission
speeds of 57,600 bps; and some modems go up to 76,800 bps.
Types of Modems
There are different types of modems because different types of communication environments require different methods of
sending data. These environments can be divided roughly into two areas related to the timing of communications:
•Asynchronous
•Synchronous

The type of modem a network uses depends on whether the environment is asynchronous or synchronous.
Asynchronous Communication (Async)
Asynchronous communication, known as "async," is possibly the most widespread form of connectivity in the world. This is
because async was developed in order to make use of common telephone lines. Figure 7.5 shows an asynchronous environment,
in which data is transmitted in a serial stream.

Each character—letter, number, or symbol—is turned into a string of bits. Each of these strings is separated from the other
strings by a start-of-character bit and a stop bit. Both the sending and receiving devices must agree on the start and stop bit
sequence. The receiving computer uses the start and stop bit markers to schedule its timing functions so it is ready to receive the
next byte of data.
Communication is not synchronized. There is no clocking device or method to coordinate transmission between the sender and
the receiver. The sending computer just sends data, and the receiving computer just receives data. The receiving computer then
checks to make sure that the received data matches what was sent. Between 20 and 27 percent of the data traffic in async
communication consists of data traffic control and coordination. The actual amount depends on the type of the transmission—
for example, whether parity (a form of error checking, discussed in the section that follows) is being used.

Asynchronous transmission over telephone lines can happen at up to 28,800 bps. However, the latest data compression methods
can boost the 28,800 bps rate to 115,200 bps over directly connected systems.

Error Control Because of the potential for error, async can include a special bit, called a parity bit, which is used in an error-
checking and correction scheme called parity checking. In parity checking, the number of bits sent must match exactly the
number of bits received.
The original V.32 modem standard did not provide error control. To help avoid generating errors during data transmission, a
company called Microcom developed its own standard for asynchronous data-error control, the Microcom Networking Protocol
(MNP). The method worked so well that other companies adopted not only the initial version of the protocol but later versions,
called classes, as well. Currently, several modem manufacturers incorporate MNP Classes 2, 3, and 4 standards.

In 1989, the Comité Consultatif Internationale de Télégraphie et Téléphonie (CCITT) published an asynchronous error-control
scheme called V.42. This hardware-implemented standard featured two error-control protocols. The primary error-control
scheme is link access procedure for modems (LAPM), but the scheme also uses MNP Class 4. The LAPM protocol is used in
communications between two modems that are V.42-compliant. If only one, but not both, of the modems is MNP 4compliant,
the correct protocol to use would be MNP 4.
Improving Transmission Performance Communication performance depends on two elements:
•Signaling or channel speed describes how fast the bits are encoded onto the communications channel.
•Throughput measures the amount of useful information going across the channel.
By removing redundant elements or empty sections, compression improves the time required to send data. Microcom's MNP
Class 5 Data Compression Protocol is an example of one current data compression standard. You can improve performance, often
doubling the throughput, by using data compression. When both ends of a communication link use the MNP Class 5 protocol,
data transmission time can be cut in half.
The V.42bis standard, because it describes how to implement impressive data compression in hardware, makes even greater
performance possible. For example, a 56.6Kbps modem using V.90 can achieve a throughput of 100Kbps.
Coordinating the Standards Asynchronous, or serial, modems are less expensive than synchronous modems because the
asynchronous modem does not need the circuitry and the components to handle the timing involved in synchronous
transmission that synchronous modems require.
Synchronous Communication
Synchronous communication relies on a timing scheme coordinated between two devices to separate groups of bits and transmit
them in blocks known as "frames." Special characters are used to begin the synchronization and check its accuracy periodically.
Because the bits are sent and received in a timed, controlled (synchronized) process, start and stop bits are not required.
Transmission stops at the end of one frame and starts again with a new one. This start-and-stop approach is much more efficient
than asynchronous transmission, especially when large packets of data are being transferred. When small packets are sent, this
increase in efficiency is less noticeable. Figure 7.6 shows a comparison of asynchronous and synchronous data streams.
If there is an error, the synchronous error-detection and correction scheme implements a retransmission.
Synchronous protocols perform a number of jobs that Synchronous communication is used in almost all digital and
asynchronous protocols do not. Principally, they: network communications. For example, if you were using digital
•Format data into blocks. lines to connect remote computers, you would use synchronous
•Add control information. modems rather than asynchronous modems to connect the
•Check the information to provide error control. computer to the digital line. Generally, their higher cost and
The primary protocols in synchronous communication are: complexity have kept synchronous modems out of the home
•Synchronous Data Link Control (SDLC). market.
•High-level Data Link Control (HDLC)
•Binary Synchronous Communications Protocol (bisync)
Asymmetric Digital Subscriber Line (ADSL)
The latest modem technology to become available is asymmetric digital subscriber line (ADSL). This technology converts existing
twisted-pair telephone lines into access paths for multimedia and high-speed data communications. These new connections can
transmit more than 8 Mbps to the subscriber and up to 1 Mbps from the subscriber.
ADSL is not without drawbacks. The technology requires special hardware, including an ADSL modem on each end of the
connection. It also requires broadband cabling, which is currently only available in a few locations, and there is a limit to the
connection length.
ADSL is recognized as a physical layer transmission protocol for unshielded twisted-pair media.

Expanding a Network Using Components

As companies grow, so do their networks. LANs tend to outgrow their original designs. You know your LAN is too small when:
•The cable begins to get crowded with network traffic.
•Print jobs include longer wait times.
•Traffic-generating applications, such as databases, experience increased response times.
The time usually comes when administrators need to expand the size or improve the performance of their networks. But
networks cannot be made larger merely by adding new computers and more cable. Each topology or architecture has limits.
There are, however, components that can be installed to increase the size of the network within its existing environment. These
components can:

•Segment existing LANs so that each segment becomes its own LAN.
•Join two separate LANs.
•Connect to other LANs and computing environments to join them into a larger comprehensive WAN.
The components that enable engineers to accomplish these goals are:
•Hubs.
•Repeaters.
•Bridges.
•Routers.
•Brouters.
•Gateways.
Hubs
Hubs can also be used to expand the size of a LAN. Although using hubs won't convert a LAN into a WAN, connecting or adding
hubs to a LAN can effectively increase the number of workstations. This method of growing a LAN is popular but does come with
many design limitations. Figure 7.7 shows how several 10BaseT hubs can be connected to expand a network.
Figure 7.8 shows how several token-ring hubs can be connected to expand a network.

NOTE: It is important to be careful when connecting hubs. Crossover cables are wired differently than standard patch cables, and
one will not work correctly in place of the other. Check with the manufacturers to determine whether you need a standard patch
cable or a crossover cable.
Repeaters
As signals travel along a cable, they degrade and become distorted in a process called "attenuation." If a cable is long enough,
attenuation will finally make a signal unrecognizable. Installing a repeater enables signals to travel farther.
How Repeaters Work
A repeater works at the physical layer of the OSI Reference Model to regenerate the network's signals and resend them out on
other segments. Figure 7.9 shows how repeaters regenerate weak signals.

The repeater takes a weak signal from one segment, regenerates it, and passes it to the next segment. To pass data through the
repeater from one segment to the next, the packets and the Logical Link Control (LLC) protocols must be identical on each
segment. A repeater will not enable communication, for example, between an 802.3 LAN (Ethernet) and an 802.5 LAN (Token
Ring).
Repeaters do not translate or filter signals. For a repeater to work, both segments that the repeater joins must use the same
access method. The two most common access methods are carrier-sense multiple-access with collision detection (CSMA/CD) and
token passing. A repeater cannot connect a segment using CSMA/CD to a segment using the token-passing access method. That
is, a repeater cannot translate an Ethernet packet into a Token Ring packet.
As shown in Figure 7.10, repeaters can move packets from one kind of physical media to another. They can take an Ethernet
packet coming from a thinnet coaxial-cable segment and pass it on to a fiber-optic segment, provided the repeater is capable of
accepting the physical connections.

Some multiport repeaters act as multiport

hubs and connect different types of media.
The same segment limits apply to networks
that use hubs, but the limits now refer to each
segment extending from a hub rather than to
the entire network.
Repeater Considerations
Repeaters afford the least expensive way to expand a network. When the need arises to extend the physical network beyond its
distance or node limitations, consider using a repeater to link segments if neither segment is generating much traffic or limiting
costs is a major consideration.
No Isolation or Filtering Repeaters send every bit of data from one cable segment to another, even if the data consists of
malformed packets or packets not destined for use on the network. This means that a problem with one segment can disrupt
every other segment. Repeaters do not act as filters to restrict the flow of problem traffic.
Repeaters will also pass a broadcast storm along from one segment to the next, back and forth along the network. A broadcast
storm occurs when so many broadcast messages are on the network that the number is approaching the network bandwidth
limit. If a device is responding to a packet that is continuously circulating on the network, or a packet is continuously attempting
to contact a system that never replies, network performance will be degraded.
Implementing a repeater This section summarizes what you need to consider when deciding whether to implement repeaters in
your network.
Use a repeater to:
•Connect two segments of similar or dissimilar media.
•Regenerate the signal to increase the distance transmitted.
•Pass all traffic in both directions.
•Connect two segments in the most cost-effective manner.

NOTE Repeaters improve performance by dividing the network into segments, thus reducing the number of computers per
segment. When using repeaters to expand a network, don't forget about the 5-4-3 rule.
Do not use a repeater when:
•There is heavy network traffic.
•Segments are using different access methods.
•Data filtering is needed.

Bridges
Like a repeater, a bridge can join segments or workgroup LANs. Figure 7.11 shows a bridge connecting two network segments.
However, a bridge can also divide a network to isolate traffic or problems. For example, if the volume of traffic from one or two
computers or a single department is flooding the network with data and slowing down the entire operation, a bridge could
isolate those computers or that department.

Bridges can be used to:

•Expand the length of a segment.
•Provide for an increased number of computers
on the network.
•Reduce traffic bottlenecks resulting from an
excessive number of attached computers.
•Split an overloaded network into two separate
networks, reducing the amount of traffic on
each segment and making each network more
efficient.
•Link unlike physical media such as twisted-pair
and coaxial Ethernet.
How Bridges Work
Because bridges work at the data-link layer of the OSI reference model, all information contained in the higher levels of the OSI
reference model is unavailable to them. Rather than distinguish between one protocol and another, bridges simply pass all
protocols along the network. All protocols pass across bridges, so it is up to the individual computers to determine which
protocols they can recognize.
The data-link layer has two sublayers: the Logical Link Control (LLC) sublayer and the Media Access Control (MAC) sublayer.
Bridges work at the MAC sublayer and are sometimes referred to as MAC-layer bridges.
A MAC-layer bridge:
•Listens to all traffic.
•Checks the source and destination addresses of each packet.
•Builds a routing table, as information becomes available.
•Forwards packets in the following manner:
• If the destination is not listed in the routing table, the bridge forwards the packets to all segments.
• If the destination is listed in the routing table, the bridge forwards the packets to that segment (unless it is the same
segment as the source).
A bridge works on the principle that each network node has its own address. A bridge forwards packets based on the address of
the destination node.
Bridges actually have some degree of intelligence in that they learn where to forward data. As traffic passes through the bridge,
information about the computer addresses is stored in the bridge's RAM. The bridge uses this RAM to build a routing table based
on source addresses.
Initially, the bridge's routing table is empty. As nodes transmit packets, the source address is copied to the routing table. With
this address information (See Figure 7.12), the bridge learns which computers are on which segment of the network.
Creating the Routing Table Bridges build their routing tables based on the addresses of computers that have transmitted data on
the network. Specifically, bridges use source addresses—the address of the device that initiates the transmission—to create a
routing table.
When the bridge receives a packet, the source address is compared to the routing table. If the source address is not there, it is
added to the table. The bridge then compares the destination address with the routing-table database.
•If the destination address is in the routing table and is on the same segment as the source address, the packet is discarded. This
filtering helps to reduce network traffic and isolate segments of the network.
•If the destination address is in the routing table and not in the same segment as the source address, the bridge forwards the
packet out of the appropriate port to reach the destination address.
•If the destination address is not in the routing table, the bridge forwards the packet to all its ports except the one on which it
originated.
In summary, if a bridge knows the location of the destination node, it forwards the packet to it. If it does not know the
destination, it forwards the packet to all segments.
Segmenting Network Traffic A bridge can segment traffic because of its routing table. As shown in Figure 7.13, a computer on
segment 1 (the source), sends data to another computer (the destination) also located in segment 1. If the destination address is
in the routing table, the bridge can determine that the destination computer is also on segment 1. Because the source and
destination computers are both on segment 1, the packet does not get forwarded across the bridge to segment 2.
Therefore, bridges can use routing tables to reduce the traffic on the network by controlling which packets get forwarded to
other segments. This controlling (or restricting) of the flow of network traffic is known as "segmenting network traffic."
A large network is not limited to one bridge. Multiple bridges can be used to combine several small networks into one large
network.
Remote Bridges
Because bridges can be such powerful tools in expanding and segmenting networks, they are often used in large networks that
have widely dispersed segments joined by telephone lines.

Only one bridge is necessary to link

two cable segments. However, where
two separate LANs are located at a
great distance from each other (See
Figure 7.14), they can be joined into a
single network. Implementing two
remote bridges connected with
synchronous modems to a dedicated,
data-grade telephone line can do this.
Because remote LAN segments can be joined over telephone lines, it is possible for multiple LANs to be joined by more than one
path. In this situation, it is possible that data might get into a continuous loop. To handle this possibility, the 802.1 Network
Management Committee of the Institute of Electrical and Electronic Engineers (IEEE) has implemented the spanning tree
algorithm (STA). Under STA, software can sense the existence of more than one route, determine which would be the most
efficient, and then configure the bridge to use that one. Other paths are disconnected using software, although the disconnected
routes can be reactivated if the primary route becomes unavailable.
Differentiating Between Bridges and Repeaters
Bridges work at a higher OSI layer than repeaters. This means that bridges have more intelligence than repeaters and can take
more data features into account.
While bridges resemble repeaters in that they can regenerate data, bridges do this at the packet level. This means that bridges
can send packets over long distances using a variety of long-distance media.
Bridge Considerations
Bridges have all of the features of repeaters, but also accommodate more nodes. They provide better network performance than
repeaters. Because bridged networks have been divided, fewer computers compete for available resources on each segment.
To look at it another way, if a large Ethernet network were divided into two segments connected by a bridge, each new network
would carry fewer packets, have fewer collisions, and operate more efficiently. Although each network would be separate, the
bridge would pass appropriate traffic between them.
Implementing a bridge
A bridge can be either a separate, stand-alone piece of equipment (an external bridge) or it can be installed in a server. If the
network operating system (NOS) supports it, one or more network interface cards (NICs), making an internal bridge, can be
installed.
Network administrators like to use bridges because they are:
•Simple to install and transparent to users.
•Flexible and adaptable.
•Relatively inexpensive.
Routers
In an environment that consists of several network segments with differing protocols and architectures, a bridge might be
inadequate for ensuring fast communication among all segments. A network this complex needs a device that not only knows the
address of each segment, but can also determine the best path for sending data and filtering broadcast traffic to the local
segment. Such a device is called a "router."
Routers work at the network layer of the OSI reference model. This means they can switch and route packets across multiple
networks. They do this by exchanging protocol-specific information between separate networks. Routers read complex network
addressing information in the packet and, because they function at a higher layer in the OSI reference model than bridges, they
have access to additional information.
Routers can provide the following functions of a bridge:
•Filtering and isolating traffic
•Connecting network segments
Routers have access to more of the information in packets than bridges have and use this information to improve packet
deliveries. Routers are used in complex networks because they provide better traffic management. Routers can share status and
routing information with one another and use this information to bypass slow or malfunctioning connections.

How Routers Work

Routers maintain their own routing tables, usually consisting of network addresses; host addresses can also be kept if the
network architecture calls for it. To determine the destination address for incoming data, the routing table includes:
•All known network addresses.
•Instructions for connection to other networks.
•The possible paths between routers.
•The costs of sending data over those paths.
As shown in Figure 7.15, a router uses its data-routing table to select the best route for the data based on costs and available
paths. Routers require specific addresses. They understand only the network numbers that allow them to communicate with
other routers and local NIC addresses. Routers do not talk to remote computers.

NOTE Remember that routing tables were also discussed in the context of bridges. The routing table maintained by a bridge
contains MAC-sublayer addresses for each node, whereas the routing table maintained by a router contains network numbers.
Although manufacturers of both types of equipment have chosen to use the term "routing table," it has a different meaning for
bridges than it does for routers.
When routers receive packets destined for a remote network, Because routers must perform complex functions on each
they send them to the router that manages the destination packet, routers are slower than most bridges. As packets are
network. In some ways this is an advantage because it means passed from router to router, data-link layer source and
routers can: destination addresses are stripped off and then re-created. This
•Segment large networks into smaller ones. enables a router to route a packet from a TCP/IP Ethernet
•Act as safety barriers between segments. network to a server on a TCP/IP Token Ring network.
•Prohibit broadcast storms, because broadcasts are not Because routers read only addressed network packets, they do
forwarded. not allow corrupted data to get passed onto the network.
Routable Protocols Not all protocols are routable. Protocols Because they do not pass corrupted data or broadcast data
that are routable include: storms, routers put little stress on networks.
•DECnet. Routers do not look at the destination node address; they look
•Internet Protocol (IP). only at the network address. Routers will pass information only
•Internetwork Packet Exchange (IPX). if the network address is known. This ability to control the data
•OSI. passing through the router reduces the amount of traffic
•Xerox Network System (XNS). between networks and allows routers to use these links more
•DDP (AppleTalk). efficiently than bridges.
Protocols that are not routable include: Using the router-addressing scheme, administrators can break
•Local Area Transport Protocol (LAT), a protocol from Digital one large network into many separate networks, and because
Equipment Corporation. routers do not pass or even handle every packet, they act as a
•NetBEUI (NetBIOS Extended User Interface). safety barrier between network segments. This can greatly
Routers are available that can accommodate multiple protocols reduce the amount of traffic on the network and the wait time
such as IP and DECnet in the same network. experienced by users.
Choosing Paths Unlike bridges, routers can accommodate multiple active paths between LAN segments and choose among
redundant paths. Because routers can link segments that use completely different data packaging and media-access schemes,
there are often several paths available for the router to use. This means that if one router does not function, the data can still be
passed over alternate routes.
A router can listen to a network and identify which parts are busiest. It uses this information to determine which path to send
data over. If one path is very busy, the router identifies an alternative path and sends data over that one.
A router decides the path the data packet will follow by determining the number of hops between internetwork segments. Like
bridges, routers build routing tables and use these in routing algorithms such as the following:
•OSPF ("open shortest path first") is a link-state routing algorithm. Link-state algorithms control the routing process and allow
routers to respond quickly to changes in the network.
•RIP (Routing Information Protocol) uses distance-vector algorithms to determine routes. Transmission Control Protocol/Internet
Protocol (TCP/IP) and IPX support RIP.
•NetWare Link Services Protocol (NLSP) is a link-state algorithm to be used with IPX.

Types of Routers
The two major types of routers are:
•Static.
•Static routers require an administrator to manually set up and configure the routing table and to specify each route.
•Dynamic.
•Dynamic routers are designed to discover routes automatically and therefore require a minimal amount of setup and
configuration. More sophisticated than static routers, they examine information from other routers and make packet-by-packet
decisions about how to send data across the network.
Table 7.2 compares and contrasts the characteristics of static and dynamic routers.

Table 7.2 Characteristics of the Two Types of Routers

Static routers Dynamic routers

Manually set up and configure all routes. Manually configure the first route. Automatically
detect additional networks and routes.
Always use the same route, determined by a routing Can choose a route based on factors such as cost
table entry. and amount of link traffic.
Use a hard-coded route (designed to handle only a Can decide to send packets over alternate routes.
specific situation), not necessarily the shortest
route.
Are considered more secure because the Can improve security by manually configuring the
administrator specifies each route. router to filter out specific network addresses and
prevent traffic from going there.
Distinguishing Between Bridges and Routers
Bridges and routers can be confusing even for engineers with LAN and WAN experience because they appear to do the same
things: both forward packets between networks and send data across WAN links.
A question often asked is how to decide when to use a bridge and when to use a router.
The bridge, which works at the MAC sublayer of the OSI data-link layer, sees only a node address. To be more specific, a bridge
looks for a node's MAC-sublayer address in each packet. If the bridge recognizes the address, it keeps the packet local or
forwards it to the appropriate segment. If the bridge does not recognize the address, it forwards the packet to all segments
except the one through which the packet arrived.

The bridge first either recognizes the

packet's MAC-sublayer address, or it
does not, and then it forwards the
packet appropriately. Figure 7.16
shows a bridge and a router and how
they relate to the OSI reference
model.
Broadcasting Forwarding the packet is the key to understanding bridges and distinguishing them from routers. With bridges,
forwarded broadcast data goes out to every computer from all ports of the bridge except the one through which the packet
arrived. That is, each computer on all networks (except the local network from which the broadcast originated) receives a
broadcast packet. In small networks this might not have much of an impact, but a large network can generate enough broadcast
traffic to slow down a network even though it is filtering for network addresses.
The router, which works at the network layer, takes more information into account than the bridge does, determining not only
what to forward but where to forward it. The router recognizes not only an address, as the bridge does, but a type of protocol as
well. Additionally, the router can identify the addresses of other routers and determine which packets to forward to which
routers.
Multiple Paths A bridge can recognize only one path
between networks. A router can search among
multiple active paths and determine which is the
best path at that particular moment.
As illustrated in Figure 7.17, if router A has a
transmission that needs to be sent to router D, it can
send the message to router C or to router B, and the
message will be forwarded to router D. Routers have
the ability to evaluate both paths and determine
which would be the best route for that transmission.
Conclusion
Four key pieces of information can help you distinguish between a bridge and a router and determine which would be
appropriate in a given situation:
•The bridge recognizes only local MAC-sublayer addresses (the addresses of NICs in its own segment). Routers recognize network
addresses.
•The bridge broadcasts (forwards) everything it does not recognize and forwards all addresses it knows, but only from the
appropriate port.
•The router works only with routable protocols.
•The router filters addresses. It forwards particular protocols to particular addresses (other routers).

Brouters
A brouter, as the name implies, combines the qualities of both a bridge and a router. A brouter can act as a router for one
protocol and as a bridge for all the others.
Brouters can:
•Route selected routable protocols.
•Bridge nonroutable protocols.
•Deliver more cost-effective and more manageable internetworking than separate bridges and routers.
Gateways
Gateways enable communication between different architectures and environments. They repackage and convert data going
from one environment to another so that each environment can understand the other environment's data. A gateway repackages
information to match the requirements of the destination system. Gateways can change the format of a message so that it
conforms to the application program at the receiving end of the transfer. For example, electronic-mail gateways, such as the
X.400 gateway, receive messages in one format, translate it, and forward it in X.400 format used by the receiver, and vice versa.
A gateway links two systems that do not use the same:
•Communication protocols.
•Data-formatting structures.
•Languages.
•Architecture.
Gateways interconnect heterogeneous networks; for example, they can connect Microsoft Windows NT Server to IBM's Systems
Network Architecture (SNA). Gateways change the format of the data to make it conform to the application program at the
receiving end.
How Gateways Work
Gateways are task-specific, which means that they are dedicated to a particular type of transfer. They are often referred to by
their task name (Windows NT Server to SNA gateway).
As shown in Figure 7.18, a gateway takes the data from one environment, strips off its old protocol stack, and repackages it in the
protocol stack from the destination network.
To process the data, the gateway:
•Disassembles incoming data through the network's complete protocol stack.
•Encapsulates the outgoing data in the complete protocol stack of the other network to allow transmission.
Some gateways use all seven layers of the OSI reference model,
but gateways typically perform protocol conversion at the
application layer. However, the level of functionality varies
widely between types of gateways.
Mainframe Gateways
One common use for gateways is to act as translators between personal computers and minicomputer or mainframe
environments. A host gateway connects LAN computers with mainframe and minicomputer systems that do not recognize
intelligent computers attached to LANs.
In a LAN environment, as shown in Figure 7.19, one computer is usually designated as the gateway computer. Special application
programs in the desktop computers access the mainframe by communicating with the mainframe environment through the
gateway computer. Users can access resources on the mainframe just as if these resources were on their own desktop
computers.
Gateway Considerations
Typically, gateways are dedicated servers on a network. They can use a significant percentage of a server's available bandwidth
because they are carrying out resource-intensive tasks such as protocol conversion. If a gateway server is used for multiple tasks,
adequate RAM and CPU bandwidth needs to be allocated or performance of the server functions will be degraded.
Gateways are considered as choices for implementation because they do not put a heavy load on internetwork communication
circuits, and they perform specific tasks efficiently.
Lesson Summary
The following points summarize the main elements of this lesson:
•Modems make it possible to communicate over telephone lines.
•There are two types of modems: synchronous and asynchronous.
•It is important to choose the right cable when connecting hubs; crossover cables will not work in place of standard patch cables.
•Asymmetric digital subscriber line (ADSL) is a technology for increasing the speed of transmission on telephone lines.
•Repeaters are used to connect two segments of similar or dissimilar media and to regenerate a signal to increase the distance
transmitted.
•Repeaters should not be used where network traffic is heavy, segments are using different access methods, or filtering is
needed.
•Bridges have all the features of repeaters.
•Bridges are used to connect two segments to expand the length or number of nodes on the network, to reduce traffic by
segmenting the network, or to connect dissimilar networks.
•Routers are used to connect two networks, limit unnecessary traffic, and to separate administrative networks.
•Brouters combine the features of bridges and routers; a brouter can act as a router for one protocol and as a bridge for all the
others.
•Gateways perform protocol and data conversion.
•Gateways are limited in several ways: they are task-specific, expensive, and can be slow.