1 Networks and the Computer Network Interface Card Presented by: Daniel D. Kochmanski Michael A. DiVito.

1

Networks and the Computer

Network Interface CardPresented by:Daniel D. KochmanskiMichael A. DiVito

2

Advantages of Networks

Allow groups of users to exchange information and share data.

Allow easy and efficient communication among individuals, including e-mail.

Allow users to share peripherals such as printers, scanners, fax machines, and other devices.

3

Local Area Networks (LANs) and Internetworks

A Local Area Network is also known as an Intranet. It works within a limited geographical area.

i.e. within one building or complex Internetwork is a network 100 or more

computers at distances in excess of 1000 feet.

4

Wide Area Networks (WANs) and the Internet

Wide Area Network (WAN) – span distances measured in miles.

i.e. two or more separate LANS linked together

Internet – global WAN internetwork; includes millions of machines and users on the world wide web connected via Network Interface Cards inside each computer.

5

Network Interface Card (NIC)

Also known as a Network Adapter. Integrated circuit board that plugs into the internal

circuitry of the computer. Allows the members of a local-area network to

communicate with each other. NIC is the physical interface from the computer or

peripheral to the medium.

The Medium may be physical cable, such as twisted pair wiring, coaxial cable, fiber optic or even wireless.

6

Network Cards Convert Data from Parallel to Serial, and vice versa

Most computers use parallel data lines internally to send data between the CPU and the adapter cards. This is called a Bus.

Most networking media transmit data in a singleline, called serial transmission.

The NIC translates parallel into serial for outgoingmessages and serial into parallel for incomingmessages.

Prior to the invention of NICs, data was sent via serial ports on the computer.

7

First Computer Networks usedSerial Communications Ports

Serial Com ports use the RS232 standardDB9 or DB25 type connector+ / - 12voltsSerial means one bit at a time.Uses

• Networking computers

• Modem

• Mouse & Printers, which are somewhat historical now

RXTX

Grnd

TXRX

Grnd

8

Asynchronous Communications

Both ends agree to a protocol and a speed. i.e. Protocol 1 start bit, 1 stop bit,

1 parity bit & 7 data bits.

Speed (baud) 300 - 115,200 bps Sender transmits start bit and both transmitting

and receiving clocks start almost simultaneously. Data is transmitted

Each clock is now working independently. The block of data is short enough so that clocks do not significantly drift out of sync.

Next block of data causes the clocks to be reset.

9

Sending a character

Line

Idle

Star

t bit

Data Bits

Pari

ty b

itSt

op b

it

Line

Idle

1 0 0 0 0 0 1

10

Receiving a character

Line

Idle

Star

t bit

Data BitsPa

rity

bit

Stop

bit

Line

Idle

Receiver’s clock

Clock

Start

Signal

Receiver reads voltage

at each agreed interval

11

Parallel Data Transmission Used on a Computer’s Bus where the adapter

card plugs in. i.e. ISA or PCI card bus slots

Used for the parallel LPT port where the printer or scanner plugs in. Parallel ports have:5Volt Logic.Transmission of 1 Byte at a time or 8 bits.Bi-directional capabilities.

Grnd Grnd

DATA

Handshake

12

Understanding the Bus inside the Computer

Bus width refers to number of parallel lines, each

able to move one bit at a time. Industry Standard Architecture (ISA) slots have an

8-bit or 16-bit bus. Peripheral Component Interconnect (PCI) slots

have a 64 bit bus and are the most popular bus used today.

13

The Network Interface Card’s Transceiver

NICs access the transceiver to transmit data onto the cable.

Most NICs include the transceiver. Some Ethernet NICs offer multiple interfaces as

you will soon see but most are designed for a specific medium.

Newer NICs generally have only one type of media connector.

14

Network Interface Card Data Packets

Packets are basic units of data for networktransmission and reception.

NICs create packages of data bits called packets, then: Transmit packets serially onto the network medium,

and Act as a gatekeeper, allowing only inbound packets

intended for its computer via an electronic serial number known as a MAC address.

15

The MAC Address

NICs have a unique identifier, called a Media Access Control (MAC) address. It is programmed into a ROM chip on the NIC. It’s a 48-bit number, written as six two-digit

hexadecimal numbers separated by colons. The first part identifies the manufacturer. The second part is unique to each NIC.

16

IP Address

Network Cards can have a second address called an IP address.

IP Address is software configurable. IP currently uses 32 bits split into four sections

separated by dots. i.e. 165.255.110.133 – These are decimal values. Only used in certain network protocols such as TCP/IP.

17

Network Protocols

Network protocols are a common set of data transmission rules that:

define how to interpret signals and identify individual computers.

initiate and end networked communication, manage information exchange across the network

medium.

Protocols include TCP/IP, NetBEUI, IPX/SPX, and NWLink

18

OSI Reference Model StructureFor Networking

Breaks networked communications intoseven layers. Layers help clarify the process of networking. Application Presentation Session Transport Network Data Link Physical

19

20

Each layer of the OSI model communicates andinteracts with layers immediately above andbelow it.

Each layer is responsible for the different aspects of data exchange.

Each layer puts an electronic envelope around dataas it sends it down the layers or removes itas it travels up the layers for delivery.

21

Data Flow

Data is broken into packets or PDUs as it moves down the stack. PDU stands for protocol data unit, packet data unit, or

payload data unit. Packets are a self-contained data structure from one

layer to another. At the sending end, each layer adds special formatting or

addressing to the packet. At the receiving end, each layer reads the packet and strips

off information added by the corresponding layer at thesending end.

22

Packet Structure

Each packet contains: Address of the target machine Address of the source machine Encapsulated Data Error Checking Data

The receiving machine checks all the packets. It accepts those with it’s address, then replies with an acknowledgement.

23

Network Interface Card Data Encapsulation

Source Serial # (48), Destination Serial # (48) and Protocol type

Source IP (32), Destination IP(32)

Offset in the byte stream, Acknowledgement, port number, connection number

HELLO THERE

Ethernet Packet IP DatagramTCP Header DATA

HELLO THERE

Trailers

24

The Basic Ethernet Frame Format contains the following seven fields

Preamble (PRE)—Consists of 7 bytes and is an alternating pattern of ones and zeros that tells the receiving stations that a frame is coming.

Start-of-frame delimiter (SOF)—Consists of 1 byte. Destination address (DA)—Consists of 6 bytes and identifies

which station should receive the frame. Source addresses (SA)—Consists of 6 bytes and it identifies

the sending station. Length/Type—Consists of 4 bytes and indicates the number

of MAC-client data bytes. Data—Is a sequence of 1500 bytes maximum. Frame check sequence (FCS)—Consists of 4 bytes and this

sequence contains a 32-bit cyclic redundancy check (CRC).

25

Basic MAC Data Frame Format

Transmission order, left to right, bit serial.

D

26

Configuring a Network Interface Card

This involves three settings: Interrupt Request line (IRQ) Base Input/Output (I/O) port Base memory address

In older computers, the user needed to supply the IRQ and base I/O port. Currently Plug and Play operating systems have automated this task and default values are normally assigned.

27

Driver Software for the NIC

This is a small specialized program that manages communications between the operating system and the NIC.

Earlier, each NIC vendor built its own driver. Now, operating system vendors define drivers for

NICs.

28

Network Layout Design

The Physical Topology refers to physical layout including computers, cables, and other resources on the network.

The Logical topology refers to how data travels between computers on the network.

29

Today’s network designs are based on three topologies:

Ring which connects computers to form a continuous circular loop.

Bus which consists of a series of computers connected along a single cable segment.

Star which connects computers via a central connection point known as a hub or switching hub.

30

Ring Topology

31

Bus Topology

32

Two Types of Coaxial Cable used as a Bus

Thin Ethernet (also called thinnet) designated by the Institute of Electrical and Electronics Engineers (IEEE) as 10Base2

Thick Ethernet (also called thicknet) designated by IEEE as 10Base5

Each of these types of cables is terminated with BNC connectors.

33

Understanding IEEE Cable Designations

10Base2, for example, refers to a total bandwidth of 10 Mbps, baseband signaling, and the maximum cable segment length as designated in hundreds of meters.

10Base2 means 200 meters including patch cables 10Base5 means 500 meters

34

Star Topology

This is the most common topology in today’s networks.

It connects computers to a central hub that receives and transmits signals to all devices on the network.

Data is addressed and sent in packets to a specific MAC address found inside the computer’s Network Interface Card.

There is ease in troubleshooting because the failure of a single computer or cable does not affect the entire network.

35

Star Topology

36

Unshielded Twisted-Pair (UTP) wire is the most common cable used in a Star topology

IEEE specifies that the most popular form of LAN cabling is 10BaseT and 100BaseT

T means Twisted-Pair wire. Normally 4 sets of twisted-pairs = 8 wires. The maximum length of a 10BaseT or a 100BaseT

segment is 100 meters or 328 feet. This is called the channel link and includes the necessary patch cable at each end of the basic link.

Twisted-Pair wire is also used for telephone systems.

37

Twisted-Pair Wire Categories

Category 1 or 2: carries voice and low speed data Category 3: has a bandwidth up to 10 Mbps. It’s

used with older networks such as 10BaseT Ethernet. It’s still commonly used for telephone networks today.

Category 4: has a bandwidth up to 16 Mbps and was formally used for 16 Mbps token ring networks.

Category 5: has a bandwidth up to 100 Mbps and is used with 100BaseT Ethernet.

Category 5e or 6: is used for Gigabit Ethernet known as 1000BaseT.

38

Twisted-Pair Connectors

Both Shielded Twisted-Pair and Unshielded Twisted-Pair wire use RJ-45 modular connectors.

These are similar to the two wire RJ-11 or four wire RJ-14 modular connectors used for the telephones in your house.

RJ-45 is larger and uses eight wires.

39

In a star topology, the network gear at the center can be a hub or a switching hub.

With a hub, data is sent from one computer to every other computer on the hub simultaneously.

While a switch, as it’s commonly called, maintains address tables for each connection.

On a switch, traffic is sent only to the port for which the data is destined.

A switch allows all pairsof stations to communicatesimultaneously at topspeed.

HUB

40

Simple Router

This router is connected to the LAN and also has a WAN (Internet) connection.

One purpose for the router is to act as a secured “gateway” protecting your network from intruders on the Internet. It uses Network Address Translation (NAT) to perform this function. NAT simply means that the router has it’s own IP address on your network and also has a uniquely different IP address on the Internet.

Router

World Wide Web

41

Types of Network Interface Cards

Fiber-Optic Network Interface Cards work on a Fiber-Optic cabled network.

Infrared networking uses infrared light to transmit data from one device to another.

Wireless network cards with antennas operate on a wireless network with a wireless hub.

Copper wired Network Interface Cards work on coax cable or twisted-pair wire.

42

10Mbs Ethernet Card with an ISA Slot

10baseT RJ45 Socket

Status LEDs

10Base2 BNCConnector

Mountingbracket

Each network cardhas a unique 48 bit identifierknown as the Media Access Control(MAC) number.

43

Electronic Components on the

Network Interface Card

Resistors Diodes Capacitors Coils Crystals Integrated Circuits such as the Realtek Fast

Ethernet controller chip shown next.

44

45M

46

MAC Layer Block Diagram

47

Features of the Ethernet Controller

Scrambling: All the encoded data is passed to the data scrambler to reduce EMI by spreading the power spectrum using a 10-bit scrambler seed loaded at the beginning.

Equalizer and Baseline Wander: High speed signals over unshielded (or shielded) twisted pair cable will experience attenuation and phase shift. These effects depend on the signal frequency, cable type, cable length and the cable connectors. Robust circuits in the transceiver provide reliable adaptive equalizer and baseline wander compensation for amplitude attenuation and phase shift due to transmission line parasitics.

4B/5B encoding procedure: Each 4-bit data nibble (representing half of a data byte) is mapped into a 5-bit binary code-group that is transmitted bit-serial over the link.

48

Signetics Thompson Ethernet Controller STE10/100A

49

Full Duplex and Half Duplex Operation of Transceiver

The transceiver can operate in either full duplex or half duplex network applications. In full duplex, both transmission and reception

between the network interface cards can take place simultaneously.

In half duplex mode, only one network card can transmit at a time.

50

System Diagram of the STE10/100A

D

51

Types of Digital to Digital Encoding

Unipolar – 1 for high, 0 for low. (common to all students)

Polar NRZ – Non-Return to Zero RZ – Return to Zero Biphase

• Manchester• Differential Manchester

Bipolar AMI – Alternate Mark Inversion, used on T1 lines B8ZS – Binary Eight Zero Suppression, used on T1 HDB3

52

Unipolar Encoding

Amplitude

Time

53

NRZ-L and NRZ-I Encoding

NRZ-L

NRZ-I

Transition cause next bit is a one.

54

RZ – Return to Zero Encoding

These transitions can be used for synchronization.

D

55

Manchester Encoding

Each bit period is divided into two equal intervals. Binary 1 = High then Low Binary 0 = Low then High Some network cards send the signal through an

inverter so this polarity is backwards.

56

Differential Manchester Encoding

Each bit period is divided into two equal intervals. Binary 1 = Lack of voltage transition from

pervious bit Binary 0 = Voltage transition from previous bit

57

Manchester Encodingvs.

Differential Manchester Encoding

58

59

Summary of the Network Interface Card

Allows connectivity of one computer to another. Controls the communication that takes place between

computers. Utilizes an ethernet controller chip to encode,

scramble, send and receive data. Converts data from parallel to serial for transmission

to another Network Interface Card. Comes in a variety of forms depending on the

application or network medium. In other words, it’s all pretty simple, huh!

60

I hope you enjoyed today’s presentation on Networks and the Network Interface Card by

Daniel D. Kochmanski and Michael A. DiVito

To view this presentation again please go to

www.ddktele.com and click on the

Networks and the Network Interface Card link.

For other information on this subject please go

to www.wzrd.com/home/ddk/nic