Chapter 6 Local Area Networks Business Data Communications

Chapter 6. Learning Objectives • Be aware of the roles of LANs in organizations • Understand the major components of LANs • Understand traditional Ethernet LANs • Understand switched Ethernet LANs • Understand wireless Ethernet LANs • Be aware of other wireless LAN technologies • Be familiar with how to improve LAN performance 2

Chapter 6. Outline • Introduction – Why use a LAN? Dedicated servers vs. Peer-to-peer LANs • LAN Components – NICs, Cables, Hubs and Network Operating Systems • Traditional Ethernet (IEEE 802. 3) – Topology, Media Access Control, Ethernet Types • Switched Ethernet – Topology, Media Access Control, Performance Benefits • Wireless LANs (IEEE 802. 11) – Topology, Media Access Control, Wireless Ethernet Types • Other Wireless Technologies – Infrared Wireless, Bluetooth • Improving LAN Performance – Improving Server Performance, Improving Circuit Capacity, Reducing Network Demand 3

Introduction 4

Why use a LAN? • There are two main benefits to using a local area network: information sharing and resource sharing. • Examples of information sharing include file sharing, exchanging e-mail, and using the Internet. • Examples of resource sharing include sharing hardware and software, such as sharing an expensive printer. • Another important resource sharing technique is to purchase software on a per seat basis. For example, only purchasing a 10 -seat license for a software program on a 20 client network instead of purchasing 20 copies of the same program. 5

Dedicated Server Networks • A basic LAN dichotomy exists between dedicated server LANs and peer-to-peer LANs which don’t have servers. Since 90% of all LANs have a dedicated server, this chapter mostly focuses on server-based LANs. • A dedicated server is a computer that is permanently assigned a specific server task such as being a Web server, e-mail server, file server or printer server. • Servers also run a special operating system called a server network operating system. • When many servers are part of a network, it can be referred to as a server farm. 6

Peer-to-Peer Networks • Peer-to-peer networks do not use dedicated servers. • Any computer on a peer-to-peer network can act as both a client, accessing resources or information on other computers on the network, or as a server, allowing access to attached information or resources. • Peer-to-peer networks tend to be small networks. • The main advantage of peer-to-peer networking is lower cost since there is no dedicated server, generally the most expensive network component. • The main disadvantage is that peer-to-peer networks are generally slower than dedicated server networks, since each computer is less powerful and may be in use as a client and a server at the same time. 7

LAN Components 8

Basic LAN Components (Figure 6 -1) • The six basic LAN components are: 1. Clients 2. Servers 3. Network Interface Cards 4. Network Cables 5. Hubs and Switches 6. Network Operating System • The first two were discussed in chapter 2, the rest will be discussed in this chapter. 9

Figure 6 -1. Basic LAN Components 10

Network Interface Cards • Network interface cards, also called network cards and network adapters include a cable socket allowing computers to be connected to the network. • NICs are part of both the physical and data link layer and include a unique data link layer address (sometimes called a MAC address), placed in them by their manufacturer. • Before sending data onto the network, the network card also organizes data into frames and then sends them out on the network. • Notebook computers often use NICs that are plugged into the PCMCIA port. 11

Network Cables • Each computer is physically connected to the network using a cable. • These cables are either untwisted wire pairs (UTP, the most common choice), shielded twisted pair (STP), coaxial cable, or optical fiber. • Wireless LANs use radio frequencies or infrared light instead of cables. • Sometimes two different types of cabling can be linked using a special connector. A BALUN (Balanced-Unbalanced) is one such device that connects UTP and Coaxial Cable. 12

Hubs • Hubs act as junction boxes, linking cables from several computers on a network. Hubs are usually sold with 4, 8, 16 or 24 ports. • Some hubs allow connection of more than one kind of cabling, such as UTP and coax. • Hubs also repeat (reconstruct and strengthen) incoming signals. This is important since all signals become weaker with distance. • The maximum LAN segment distance for a cable can therefore be extended using hubs. 13

Figure 6 -2. Network Hub 14

Network Operating Systems • The NOS is the software that runs the LAN. It comes in two types: Server NOSs & Client NOSs. • Server NOSs enable server to execute and respond to the requests sent to them as web server, print servers, file servers, etc. • Client NOS functions are typically included in most OS packages such as Windows 98 and Windows 2000. 15

Network Profiles • The network profile specifies what resources on each server are available to the network for use by other computers, including data files, printers, etc. • Devices that are not included in the network profile can not be used over the network. • User profiles describe what each user on a LAN has access to. • Most LANs also use auditing software which keeps track of which user has accessed what network resource. 16

Traditional Ethernet (IEEE 802. 3) 17

Ethernet (IEEE 802. 3) • Almost all LANs today use Ethernet • Originally, Ethernet was jointly developed by a consortium of Digital Equipment Corp. , Intel and Xerox and was standardized as IEEE 802. 3. • Ethernet LANs that use hubs are sometimes called shared Ethernet. 18

Shared Ethernet Topology (Figure 6 -3) • Ethernet’s logical topology is a bus topology. • This means all computers on the network receive messages from all other computers, whether the message is intended for those computers or not. • When a frame is received by a computer, the first task is to read the frame’s destination address to see if the message is meant for it or not. • Although, a decade ago most Ethernet LANs used a physical bus, almost all Ethernets today use a physical star topology, with the network’s computers linked into hubs. • It is also common to link use multiple hubs to form more complex physical topologies (Figure 6 -4). 19

Figure 6 -3 Ethernet Topology 20

Figure 6 -4 Multiple Hub Ethernet Design 21

Media Access Control • Ethernet’s medium access control protocol, called CSMA/CD, is contention-based • With a contention-based protocol, frames can be sent by two computers on the same network at the same time, in which case they will collide and become garbled. • CSMA/CD, can thus be termed “ordered chaos” because it tolerates, rather than avoids, collisions caused by two computers transmitting at the same time. 22

CSMA/CD • Stands for: Carrier Sense Multiple Access w/ Collision Detect • Carrier Sense: computers listen to the network to see if another computer is transmitting before sending anything themselves. • Multiple Access: all computers have access to the network medium. • Collision Detect: if they detect a collision (CD), they then wait a random amount of time and resend the frame (It has to be random in order to avoid another collision). 23

Ethernet Physical Media Standards • Ethernet Media are formatted as follows: [Value 1]Base/Broad[Value 2] • Value 1: Data Rate for Medium 10 = 10 Mbps • Base or Broad – Base = Baseband Mode meaning only one (digital) channel – Broad = Broadband (analog) cable transmissions use more than one channel (e. g. , cable TV) • Value 2: (relates to maximum distance possible in hundreds of meters or cable type T= twisted pair, F =fiber) 24

Types of Ethernet (Figure 6 -5) • Seven types of shared Ethernet have been in use: • 10 Base 5 = thick Ethernet, uses thick coax. This is the original Ethernet specification. Now uncommon. • 10 Base 2 = thin Ethernet, uses thin coax. Became popular in the early 1990 s as a cheaper alternative to 10 Base 5. Now uncommon. • 10 Base. T = twisted pair Ethernet, most common type of Ethernet. Uses Cat 3 and Cat 5 UTP. Common but rapidly losing ground to 100 Base. T. • 100 Base. T = also called Fast Ethernet, has replaced 10 Base. T in sales volume. Uses Cat 5 UTP (Sometimes combined 10/100 Ethernet is found in which some segments run 10 Base. T and some run 100 Base. T is also used by some organizations). • 1000 Base. T = Gigabit Ethernet. Maximum cable length is only 100 meters. • 10 Gb. E = 10 Gbps Ethernet. Uses fiber and is typically full duplex. • 40 Gb. E = 40 Gbps Ethernet. Uses fiber and is typically full duplex. 25

Name Maximum Data Rate Cables 10 Base 5 10 Mbps Coaxial 10 Base 2 10 Mbps Coaxial 10 Base. T 10 Mbps UTP cat 3, UTP cat 5 100 Base. T 100 Mbps UTP cat 5, fiber 1000 Base. X 1 Gbps UTP cat 5, UTP cat 5 e, UTP cat 6, fiber 10 Gb. E 10 Gbps UTP cat 5 e, UTP cat 6, UTP cat 7, fiber 40 Gb. E 40 Gbps fiber Figure 6 -5 Types of Ethernet 26

Switched Ethernet 27

Switched Ethernet Topology • Switched Ethernet uses switches instead of hubs. • While a hub broadcasts frames to all ports, the switch reads the destination address of the frame and only sends it to the corresponding port. • The effect is to turn the network into a group of point-to-point circuits and to change the logical topology of the network from a bus to a star. 28

Basic Switch Operation • Switches make forwarding decisions based on forwarding tables (similar to routing tables). • When a frame is received, the switch reads its [data link layer] destination address and sends the frame out the corresponding port in its forwarding table. • Switches making switching decisions based on data link layer addresses are called layer-2 switches. • When a switch is first turned on, its forwarding table is empty. It then learns which ports correspond to which computers by reading the source addresses of the incoming frames along with the port number that the frame arrived on. • If the switch’s forwarding table does not have the destination address of the frame, it broadcasts the frame to all ports. • Thus, a switch starts by working like a hub and then works more and more as a switch as it fills its forwarding table. 29

Media Access Control • Switched Ethernet still uses CSMA/CD media access control, but collisions are less likely as each network segment operates independently. • The network’s modified topology also allows multiple messages to be sent at one time. • For example, computer A can send a message to computer B at the same time that computer C sends one to computer D. • If two computers send frames to the same destination at the same time, the switch stores the second frame in memory until it has finished sending the first, then forwards the second. 30

Figure 6 -6 802. 3 Ethernet versus switched Ethernet 31

Performance Benefits • Switched Ethernet can dramatically improve network performance. • Shared Ethernet 10 Base. T networks are only capable of using about 50% of capacity before collisions are a problem • Switched Ethernet, however, runs at up to 95% capacity on 10 Base. T. • Another performance improvement can be made by using a 10/100 switch that uses a 100 Base. T connection for the server(s) and/or routers, i. e. , the network segments experiencing the highest volume of LAN traffic. 32

Wireless Ethernet (IEEE 802. 11) 33

Wireless Ethernet (IEEE 802. 11) • Wireless LANs dispense with cables and use radio or infrared frequencies to transmit signals through the air. • WLANs are growing in popularity because they eliminate cabling and facilitate network access from a variety of locations and for mobile workers (as in a hospital). • The most common wireless networking standard is IEEE 802. 11, often called Wireless Ethernet or Wireless LAN. 34

Wireless LAN Topology • WLAN topologies are the same as on Ethernet: physical star, logical bus (Figure 6 -7). • Wireless LAN devices use the same radio frequencies, so they must take turns using the network. • Instead of hubs, WLANs use devices called access points (AP). Maximum transmission range is about 100 -500 feet. Usually a set of APs are installed making wireless access possible in several areas in a building or corporate campus. • Each WLAN computer uses an NIC that transmits radio signals to the AP. • Because of the ease of access, security is a potential problem, so IEEE 802. 11 uses 40 -bit data encryption to prevent eavesdropping. 35

Figure 6 -7 A wireless Ethernet access point connected into an Ethernet Switch. 36

WLAN Media Access Control • Wireless LANs use CSMA/CA where CA = collision avoidance (CA). With CA, a station waits until another station is finished transmitting plus an additional random period of time before sending anything. • Two different WLAN MAC techniques are now in use: the Physical Carrier Sense Method and the Virtual Carrier Sense Method. 37

Physical Carrier Sense Method • In the physical carrier sense method, a node that wants to send first listens to make sure that the transmitting node has finished, then waits a period of time longer. • Each frame is sent using the Stop and Wait ARQ, so by waiting, the listening node can detect that the sending node has finished and can then begin sending its transmission. • With Wireless LANs, ACK/NAK signals are sent a short time after a frame is received, while stations wishing to send a frame wait a somewhat longer time, ensuring that no collision will occur. 38

Virtual Carrier Sense Method • When a computer on a Wireless LAN is near the transmission limits of the AP at one end another computer is near the transmission limits at the other end of the AP’s range, both computers may be able to transmit to the AP, but can not detect each other’s signals. • This is known as the hidden node problem. When it occurs, the physical carrier sense method will not work. • The virtual carrier sense method solves this problem by having a transmitting station first send a request to send (RTS) signal to the AP. If the AP responds with a clear to send (CTS) signal, the computer wishing to send a frame can then begin transmitting. 39

Types of Wireless Ethernet • Two forms of the IEEE 802. 11 b standard currently exist: • Direct Sequence Spread Spectrum (DSSS) uses the entire frequency band to transmit information. DSSS is capable of data rates of up to 11 Mbps with fallback rates of 5. 5, 2 and 1 Mbps. Lower rates are used when interference or congestion occurs. • Frequency Hopping Spread Spectrum (FHSS) divides the frequency band into a series of channels and then changes its frequency channel about every half a second, using a pseudorandom sequence. FHSS is more secure, but is only capable of data rates of 1 or 2 Mbps. • IEEE 802. 11 a is another Wireless LAN standard that is still being defined. It will operate in the 5 GHz band be capable of data rates of up to 54 Mbps, but will probably average about 20 Mbps in practice. 40

Other Wireless Technologies 41

Infrared Wireless LANs • Infrared WLANs are less flexible than IEEE 802. 11 WLANs because, as with TV remote controls that are also infrared based, they require line of sight to work. • Infrared Hubs and NICs are usually mounted in fixed positions to ensure they will hit their targets. • The main advantage of infrared WLANs is reduced wiring. • A new version, called diffuse infrared, operates without a direct line of sight by bouncing the infrared signal off of walls, but is only able to operate within a single room and at distances of only about 50 -75 feet. 42

Fig. 6 -9 Infrared Wireless LAN 43

Bluetooth • Bluetooth is a 1 Mbps wireless standard developed for piconets, small personal or home networks. • It may soon be standardized as IEEE 802. 15. • Although Bluetooth uses the same 2. 4 GHz band as Wireless LANs it is not compatible with the IEEE 802. 11 standard and so can not be used in locations that use the Wireless LANs. • Bluetooth’s controlled MAC technique uses a master device that polls up to 8 “slave” devices. • Examples of Bluetooth applications include; linking a wireless mouse, a telephone headset, or a Palm handheld computer to a home network. 44

Improving LAN Performance 45

Improving LAN Performance • As networks become more and more intensively used, LAN performance becomes a critical issue. • The measure of LAN Performance is throughput, i. e. , the total amount of user data transmitted in a given period of time. • LAN performance can be improved by identifying and eliminating bottlenecks, that is, points in the network where congestion is occurring because the network or device can’t handle all of the demand it is experiencing. 46

Identifying Network Bottlenecks • Two common network bottlenecks are related to server access: • If server performance is poor when server utilization is high (>60%), then the bottleneck is the server. • If server performance is poor during periods of low server utilization (<40%), then the bottleneck is not the server but the network circuit. 47

Improving Server Performance • Two types of server performance improvements are possible: – Software improvements such as choosing a faster Network Operating System, fine tuning network and NOS parameters for optimal performance. – Hardware improvements such as adding a second server, upgrading the server’s CPU, increasing its memory space, adding more hard drives or adding a second NIC to the server. 48

Improving Server Performance: RAID • Improving disk drive performance is especially important, since disk reads are the slowest task the server needs to do. • Replacing one large drive with many small ones can improve server performance. • RAID or Redundant Array of Inexpensive Disks, builds on this idea. RAID system can be used to both improve performance and increase reliability by building redundancy into the hard drives, so that a hard drive failure does not result in any loss of data. 49

Improving Circuit Capacity • Improving circuit capacity can be done simply by upgrading one or all segments of a network to a faster protocol (which also means upgrading the NICs), such as; – Upgrading the network from 10 Base. T to 100 Base. T, or – Upgrading the network segment to the server from 10 Base. T to 100 Base. T • Another approach to improving circuit capacity is by increasing the number of network segments to the server. Most servers can handle several network segments simply by adding additional NIC cards, thereby increasing access to the server (see Figure 6 -11). 50

Fig. 6 -11 b Network Segmentation: a. Before b. After 51

Reducing Network Demand • Performance can also be improved by reducing network demand. This can be done by: – Moving more files, such as heavily used software packages to client computers. – Disk caching, software on client machines can also reduce server demand. – Moving user demands from peak times to off peak times, by telling network users when peak usage times occur and encouraging users to not use the network as heavily during these times can also help improve performance. – Delaying some network intensive jobs to off-peak times, such as running heavy printing jobs at night, can also improve performance. 52

Figure 6 -10 Improving LAN performance • Increase Server Performance – Software: Fine-tune the NOS settings – Hardware: • Add more servers and spread the network applications across the servers to balance the load • Upgrade to a faster computer • Increase the server's memory • Increase the number and speed of the server's hard disk(s) • Upgrade to a faster NIC • Increase Circuit Capacity – Upgrade to a faster circuit – Segment the network • Reduce Network Demand – Move files from the server to the client computers – Increase the use of disk caching on client computers – Change user behavior 53

End of Chapter 6 54