Network Devices: ENG224 Information Technology - Part I
Network Devices: ENG224 Information Technology - Part I
Network Devices: ENG224 Information Technology - Part I
8. Network Devices
8. Network Devices
1
A. Expanding Network
Networks cannot be made larger by simply adding new computers and more cables Less efficient !! Can install components to segment (divide) large LAN to form smaller LANs connect LANs Required components Repeaters, bridges, routers, brouters, switches or gateways
b. Bridges
Has one input and one output Used to isolate network traffic and computers Has the intelligent to examine incoming packet source and destination addresses But cannot interpret higherlevel information Hence cannot filter packet according to its protocol
6
Bridges work at the Media Access Control Sublayer of the OSI model Routing table is built to record the segment no. of address If destination address is in the same segment as the source address, stop transmit Otherwise, forward to the other segment 7
Creating a Switching Table Based on the addresses of the sending computers New addresses are added if they are not in the table Add02 S 02 D 01
Add01
Stop
01 02
Add03
S 01
D 02
Remote Bridges Bridges are often used in large networks that have widely dispersed segments Remote bridges can be used to connect remote segments via data-grade telephone line
10
c. Switches
Switches operate at the Data Link layer (layer 2) of the OSI model Can interpret address information Switches resemble bridges and can be considered as multiport bridges By having multiports, can better use limited bandwidth and prove more cost-effective than bridge
11
Cisco Catalyst 2900 switch
Switches divide a network into several isolated channels Packets sending from 1 channel will not go to another if not specify Each channel has its own capacity and need not be shared with other channels
Hub 3.3Mbps
10Mbps 3.3Mbps
Switch
10Mbps 10Mbps
3.3Mbps
12
10Mbps
Advantages of Switches Switches divide a network into several isolated channels (or collision domains) Reduce the possibility of collision
Collision only occurs when two devices try to get access to one channel Can be solved by buffering one of them for later access
Limitations of Switches Although contains buffers to accommodate bursts of traffic, can become overwhelmed by heavy traffic Device cannot detect collision when buffer full
CSMA/CD scheme will not work since the data channels are isolated, not the case as in Ethernet
Those data packets are continuously pumped to the switch and introduce more problems
14
2/6 Bytes
Read the first 14 bytes of each packet, then transmit Much faster Cannot detect corrupt packets Can propagate the corrupt packets to the network Best suited to small workgroups
15
Method of Switching - Store and Forward Mode Read the whole packet before transmit Slower than the cut-through mode More accurate since corrupt packets can be detected using the FCS More suit to large LAN since they will not propagate error packets Facilitate data transfer between segments of different speed
16
DB 100Mbps
10Mbps
Using Switches to Create VLANs Switches can logically group together some ports to form a virtual local area network (VLAN)
SW1 VLAN1 VLAN2
SW2
Hub
Hub
SW3
Switches can be configured to communicate only within the devices in the group Hub
17
d. Routers
Layer 2 Switches cannot take advantage of multiple paths Routers work at the OSI layer 3 (network layer) They use the logical address of packets and routing tables to determine the best path for data delivery
18
How Routers Work As packets are passed from routers to routers, Data Link layer source and destination addresses are stripped off and then recreated Enables a router to route a packet from a TCP/IP Ethernet network to a TCP/IP token ring network Only packets with known network addresses will be passed - hence reduce traffic Routers can listen to a network and identify its busiest part Will select the most cost effective path for 19 transmitting packets
How Routing Table is formed Routing table is formed based on communications between routers using Routing Protocols Routing Protocols Routable Protocol Routing Protocols collect data about current network status and contribute to selection of the best path
20
Routers communicate within themselves
Routing Protocol Example - RIP for IP Routing RIP (Routing Information Protocol) the oldest one Use no. of hops between nodes to determine best path Does not consider the network congestion condition Broadcast every 30 sec the routing table to neighbouring routers to convey routing information RIP is limited to interpreting a maximum of 16 hops Not suitable for large network (e.g. Internet) Can create excessive network traffic due to broadcasting
21 May take a long time to reach the far reaches
If link failure notice is received, router can rapidly compute an alternate path
22
Manual configuration Manual configuration of the first of routes route. Automatic discovery of new routes Always use the same Can select the best route route More secure Need manual configuration to improve security
23
Layer-3 Switches
Layer-3 switches operate in both layer 2 (data link layer) and 3 (network layer) Can perform both MAC switching and IP routing A combination of switch and router but much faster and easier to configure than router
25
Summary
Repeaters are the least expensive way to expand a network, but they are limited to connecting two segments Bridges function similar to repeaters, but can understand the node addresses Switches can be considered as multiport bridges, can divide a network into some logical channels Routers interconnect networks and provide filtering functions. They can determine the best route
26
1. Modems
Allow computers to communicate over a telephone line Enable communication between networks or connecting to the world beyond the LAN
27
Cannot send digital signal directly to telephone line Sending end: MODulate the computers digital signal into analog signal and transmits Receiving end: DEModulate the analog signal back into digital form
28
1 Amplitude Modulation
Frequency Modulation
Phase Modulation Normal sine wave
29
Modems typically have the following I/O interface: A serial RS-232 communication interface An RJ-11 telephone-line interface (a telephone plug)
30
RS-232
RJ-11
Modem Standards
Standard V.22bis V.32 V.32bis V.32terbo bps Introduced 2,400 1984 9,600 1984 14,400 1991 19,200 1993 1993 1994 1995 1998 Remarks
Communicate only with another V.32terbo (V.FC) Improved V.FC With compression Resolved competition between X2 and Flex56k
31
Baud rate - the number of symbol change per second on the transmission line Bit per second (bps) - number of bits transmitted per second In the past, they are identical With compression technique, a change of signal can mean more than one bits 28.8kbaud can mean 115.2kbps when using 32 V.42bis
How V.90 Works Modem speed is determined by channel noise level The noise level of traditional PSTN (public switch telephone network) limits data rate to ~35kbps 56K modem technology assumes only one analog link hence noise level is much lower
33
Why V.90 cannot achieve 56kbps in practice? The actual data link is 64kbps To prevent interference and allow some overhead data in communication, ITU recommends a lower rate to 56 kbps However, 56 kbps is a theoretical number Depending on the quality and length of the analog link, the actual data rate can range from 30kbps to 53kbps
34
No clocking devices Commonly used in telephone networks Data is transmitted in a serial stream. Each character is turned into a string of 8 bits Each of these characters is separated by one start bit and one or two stop bits
35
Types of Modem - Synchronous Modems Need clocking devices Data are transmitted in blocks Used in digital networks
36
Comparison Asynchronous modems are relatively simple and economic Large overhead - can be up to 20 to 27% of the data traffic Error control is done by using parity bit or higher layer protocols, e.g. MNP, V.42 Synchronous modems are relatively complicated and expensive Seldom use in home market Less overhead means higher efficiency More sophisticated error control protocol is 37 required
2. ADSL
ADSL stands for Asymmetric Digital Subscriber Line Particularly suitable for high speed multimedia communications, general Internet applications Asymmetric - downstream 1.5 to 6.1Mbps upstream 16 to 640kbps Digital - mainly for transmitting digital data still require modulation and demodulation Subscriber line - make use of the analog connection between household and CO
38
normal voice
ADSL Illustration
2 to 3 miles
subscriber line
Splitter local loop Telephone Company
low speed
data high speed
39
Why Asymmetric? In general Internet applications, downstream often requires a higher data rate than upstream Downstream - file download, video playback Upstream - click a link, send a form Reducing the resource for upstream can provide more resource for downstream
40
Why Subscriber Line? By better controlling the length and quality of the analog connection between household and CO, a higher data rate can be achieved
Data Rate Wire Gauge Distance Wire Size 1.5 or 2 Mbps 24 AWG 18,000 ft 0.5 mm 1.5 or 2 Mbps 26 AWG 15,000 ft 0.4 mm 6.1 Mbps 24 AWG 12,000 ft 0.5 mm 6.1 Mbps 26 AWG 9,000 ft 0.4 mm Distance 5.5 km 4.6 km 3.7 km 2.7 km
41
More than 80% of the current installed subscriber lines can fulfill this requirement Hence no extra cabling is required
ISP POP
Internet
Firewall Router DHCP DNS SNMP
DSLAM
Modem
ADSL Modem
N I D
Copper Loop
T N I
DSLAM - Digital subscriber line access module (central 42 office ADSL modem pool)
HDSL High speed DSL 2 twisted pair, 12,000 feet 1.5Mbps (DS1) full-duplex Symmetric VDSL Very high bit rate DSL Downstream: 52 Mbps (SONET STS-1) over 1000 feet; or 15 Mbps over 3000 feet Upstream: 1.5 to 2.3 Mbps RDSL Rate adaptive DSL Intelligent DSL to adjust data rate 43