WEB

TECHNOLOGY
UTTAM K. ROY

Dept. of Information Technology,

Jadavpur University, Kolkata
 Web technology 2

Agenda

 Background
 HTTP Protocol
 Domain Name System (DNS)
 Simple Mail Transfer Protocol (SMTP)
 HyperText Markup Language(HTML)
 JavaScript
 XML
 JSP

| U. K. Roy | [ 10/22/08 ]
HyperText
Tr ansf er
Pr otocol
(H TTP )
 Web technology 4

WWW
• World Wide Web—a repository of Information
• Introduced in 1991
• Originated from the CERN High-Energy Physics
laboratory in Geneva, Switzerland.
• Purpose—create a system to handle distributed resource
• A client-server service
• Service provider—called website

| U. K. Roy | [ 10/22/08 ]
 Web technology 5

The Web: Some Jargon

• Web page
– consists of objects (HTML file, JPEG image, GIF image…)
– addressed by URL
• Most Web pages consist of
– base HTML page
– several referenced objects—Hypertext and Hepermedia
• URL
– A standard way of specifying the location of an object, typically
a web page, on the Internet
• User agent for Web is called a browser
– Windows
• MS Internet Explorer
– Linux
• Netscape Navigator
• Mozzila
• Konquor
• Server for Web is called a Web server
| U. K. Roy | [ 10/22/08 ]
 Web technology 6

HyperText Transfer Protocol

• Web’s application layer protocol
– Used to access data on the World Wide Web
– Rapid jump from one document to another
• Client-server model
– client: browser that requests, receives, “displays” web
objects
– server: Web server sends objects in response to request
• uses TCP connection on the well-known port 80

| U. K. Roy | [ 10/22/08 ]
 Web technology 7

URL

• An address of the web page or other information on

the Internet
• Example
– http://www.yahoo.com/
– http://www.jusl.ac.in/images/sitemap.gif
– http://www.foldoc.org/?Uniform+Resource+Locator
– http://mail.jusl.ac.in/
– http://www.itd.jusl.ac.in:8080/jsp/test.jsp
– ftp://wuarchive.wustl.edu/mirrors/msdos/graphics/gifkit.zip
| U. K. Roy | [ 10/22/08 ]
 Web technology 8

URL - continued

• Method
– protocol used to retrieve the document (FTP, HTTP, …)
• Host
– a computer where the info is located
– the name/IP address of the computer can be an alias (not
necessary www)
• Port
– optional port # of the server (default is 80)
• Path
– the path name of the file where the info is located

| U. K. Roy | [ 10/22/08 ]
 Web technology 9

HTTP - example
• Suppose user enters URL www.yahoo.com/index.html

1a. http client initiates TCP connection to

http server (process) at www.yahoo.com.
Port 80 is the default for http server

1b. http server at host www.yahoo.com

waiting for TCP connection at port 80
“accepts” connection, notifying client

2. http client sends http request message

(containing URL) into TCP connection
socket

3. http server receives request message,

forms response message containing
requested object (index.html), sends
time
message into socket

| U. K. Roy | [ 10/22/08 ]
 Web technology 10

HTTP – example (cnt’d)

4. http server closes TCP connection

5. http client receives response message

containing html file, parses html file (using
browser), finds embedded image, and
finally displays in the browser

6. steps 1-5 repeated for another resource

time

| U. K. Roy | [ 10/22/08 ]
 Web technology 11

HTTP protocol – message format

• two types of messages: request & response
• HTTP request message
HTTP/0.9
HTTP/1.0
HTTP/1.1

GET – when the client wants to retrieve a document

from the server
HEAD – when the client wants some info about a
document but not document itself
COPY – copies the file to another location

| U. K. Roy | [ 10/22/08 ]
 Web technology 12

Other Request type (method)

Method Description
POST Used to provide information (e.g. input) to the server
Used to provide a new or replacement document to be
PUT
stored on the server
Similar to PUT except that the request contains only list of
PATCH
differences that should be implemented in the existing file
MOVE Used to copy a file to another location
DELETE Used to remove a document from the server
Used to create a link or links of a document to another
LINK
location
UNLINK Used to delete link created by LINK
OPTION Used by the client to ask the server about abailable options

| U. K. Roy | [ 10/22/08 ]
 Web technology 13

HTTP – message format

• HTTP response message
http://www.w3.org/Protocols/HTTP/HTRESP.html

explains the status code

in text form

200 OK – request succeeded

301 Moved Permanently – object moved
400 Bad Request – not understood by server
404 Not Found – req. document not found

| U. K. Roy | [ 10/22/08 ]
 Web technology 14

HTTP – message format (Status code)

100 range Informational

200 range Successful request
300 range Redirectional
400 range Client Error
500 range Server Error

| U. K. Roy | [ 10/22/08 ]
 Web technology 15

HTTP – message format (Status code)

Code Phrase Description
Informational
100 Continue The initial part of the request has been received and
the client may continue with its request
101 Switching The server is complying with a client request to
switch protocols defined in the upgrade header
Success
200 OK The request is successful
201 Created A new URL is created
202 Accepted The request is accepted, but it is not immediately
acted upon
204 No content There is no content in the body
Redirection
301 Multiple choices The requested URL refers to more than one request
302 Moved permanently The requested URL is no longer used by the server
304 Moved temporarily The requested URL has moved temporarily
| U. K. Roy | [ 10/22/08 ]
 Web technology 16

HTTP – message format (Status code)

Code Phrase Description

Client Error
400 Bad Request There is a syntax error in the request
401 Unauthorized The request lacks proper authorization
403 Forbidden Service is denied
404 Not found The document is not found
405 Method not allowed The method is not supported in this URL
406 Not acceptable The format request is not acceptable
Server Error
500 Internal Server There is an error, such as crash, the server side
Error
501 Not Implemented The action requested can not be performed
503 Service The service is temporarily unavailable, but may be
unavailable requested in the future

| U. K. Roy | [ 10/22/08 ]
 Web technology 17

HTTP – message format

• HTTP response message

| U. K. Roy | [ 10/22/08 ]
 Web technology 18

HTTP – message format

• Headers
– exchange additional
information between the
client & the server
– example
• Date
• Client’s email address
• Document age
• Content length

| U. K. Roy | [ 10/22/08 ]
 Web technology 19

HTTP – message format

General Header

Header Description
Cache-control Specifies information about caching
Shows whether the connection should be closed or
Connection
not
Date Shows the current date
MIME-version Shows the MIME version used
Upgrade Specifies the preferred communication protocol

| U. K. Roy | [ 10/22/08 ]
 Web technology 20

HTTP – message format (Request Header)

Header Description
Accept Shows media format the client can accept
Accept-charset Shows the character set the client can handle
Accept-encoding Shows the encoding scheme the client can handle
Accept-language Shows the language the client can accept
Authorization Shows the permission the client has
From Shows the email address of the user
Host Shows the host and port number of the client
If-modified-since Send the document if newer than specified date
If-match Send the document only if matches given tag
If-non-match Send the document only if does not match given tag
If-range Send only the portion of the document that is missing
If-unmodified-
Send the document if not changed since specified date
since
Referrer Specifies the URL of the linked document
User-agent Identifies the client program

| U. K. Roy | [ 10/22/08 ]
 Web technology 21

HTTP – message format (Response Header)

• Specifies the server’s configuration and special information

about the request

Header Description
Shows if server accepts the range requested by
Accept-range
client
Age Shows the age of the document
Public Shows the supported list of methods
Specifies the date after which the server will be
Retry-after
available
Server Shows the server name and version number

| U. K. Roy | [ 10/22/08 ]
 Web technology 22

HTTP – message format (Entity Header)

• Specifies information about the body
Header Description
Allow List of valid methods that can be used with a URL
Content-encoding Specifies the encoding scheme
Content-language Specifies the language
Content-length Shows the length of the document
Content-range Specifies the range of the document
Content-type Specifies the media type
Etag Gives an entity tag
Gives the date and time when contents may
Expires
change
Last-modified Gives the date and time of the last change
Specifies the location of the created of moved
Location
document
| U. K. Roy | [ 10/22/08 ]
 Web technology 23

HTTP messages – an example

This example retrieves a document.

We use the GET method to retrieve an image with the path /usr/bin/image1.
The request line shows the method (GET), the URL, and the HTTP version
(1.1).
The header has two lines that show that the client can accept images in GIF
and JPEG format.
| U. K. Roy | [ 10/22/08 ]
 Web technology 24

HTTP messages – an example

This example retrieves information about a document. We use the HEAD

method to retrieve information about an HTML document

| U. K. Roy | [ 10/22/08 ]
 Web technology 25

Persistent and nonpersistent connections

• Nonpersistent • Persistent
– HTTP 1.0 – default for HTTP 1.1
– one TCP connection for – the server leaves the TCP
each request/response connection open for more
3. the client opens a TCP requests after sending a
connection and sends a response
request – client sends requests for
4. the server sends the all referenced objects as
response and closes the soon as it receives base
connection HTML
5. the client reads data and • pipelining
closes the connection – fewer RTT

– each object transfer is

independent

| U. K. Roy | [ 10/22/08 ]
 Web technology 26

Web caches - Proxy

• HTTP supports Proxy servers
• Proxy server
1. a computer that keeps copies of responses to recent requests
• Goal: satisfy client request without involving original server
Origin
server
• client sends all http requests to the
http Proxy
proxy server
response server
client
• if object at web cache sends the object
in http response http
request http
• else request object from the origin response
server, then returns http response to
http
client request

client Origin
server

| U. K. Roy | [ 10/22/08 ]
 Web technology 27

Why Web caching?

• Assume: cache is close to
a client (in the same
network) the Internet
– smaller response
time (improved
latency)
1.544 Mbps link
– decrease traffic to
distance servers
institutional
• link out of ISP network 10 Mbps LAN
network
is often a bottleneck

institutional
cache

| U. K. Roy | [ 10/22/08 ]
 Web technology 28

Consistency of Web caching

• The major issue: How to maintain consistency?
• Two ways:
– Pull
• Web caches periodically pull the web server to see if a
document is modified
– Push
• Whenever a server gives a copy of a web page to a web
cache, they sign a lease with an expiration time;
if the web page is modified before the lease, the server
notifies the cache

| U. K. Roy | [ 10/22/08 ]
Domain Name
System
(DNS)
 Web technology 30

Domain Name System (DNS)

• TCP/IP uses IP address—difficult to remember

• Solution: use names instead of IP addresses

• Used to map a name to an IP address & vice-

versa
– example:
• www.itd.jusl.ac.in -> 203.197.107.107
• www.yahoo.com -> 209.73.186.238
• www.google.com -> 64.233.189.104

| U. K. Roy | [ 10/22/08 ]
 Web technology 31

Domain Name System (DNS)

• Possible solution:
– a host file, two columns: name & address
– Every host stores this file
– Update periodically from master file
• Problems:
– Host file would be too large to store
– Updation problem
– Solution
• Store this host file centrally
• Problem: Huge amount of traffic

| U. K. Roy | [ 10/22/08 ]
 Web technology 32

Domain Name System (DNS)

• Solution for huge amount of information:

– divide it into smaller parts and store each part on
different computer—called DNS Server
– Host needs name resolution contacts nearest
DNS Server
– if one DNS server doesn't know how to translate
a particular domain name, it asks another one,
and so on, until the correct IP address is
returned.
| U. K. Roy | [ 10/22/08 ]
 Web technology 33

Domain Name System (DNS)

• Name space
– flat name space
• Centrally controlled to avoid ambiguity and duplication
– cannot be used in larger networks like the Internet

– hierarchical name space

• each name is made of several parts
• central authority only partially control names
(www.jadavpur.edu)
– www.itd.jusl.ac.in
– www.cse.iitk.ac.in

| U. K. Roy | [ 10/22/08 ]
 Web technology 34

Domain Name Space

• designed to have a hierarchical name space
• tree structure (maximum 128 levels)

label

• all labels (maximum of 63 characters) have different names

• uniqueness of the domain names
• root label - null

| U. K. Roy | [ 10/22/08 ]
 Web technology 35

Domain name
• Domain name – a sequence of labels separated by dots
• read from the node up to the root
• full domain name ends with the null

• Fully qualified domain name

challenger.atc.fhda.edu.

• Partially qualified domain name

challenger

| U. K. Roy | [ 10/22/08 ]
 Web technology 36

Domain
• A sub-tree of the Domain Name Space
• Name of a domain is the domain name of the node at
the root of the subtree

| U. K. Roy | [ 10/22/08 ]
 Web technology 37

Distribution of name space

• Storing all naming information in one computer is
– unreliable
– inefficient
• Responding to requests from all over the world places a
heavy load on the system
• Hierarchy of Name Servers

| U. K. Roy | [ 10/22/08 ]
 Web technology 38

DNS zones, servers -

• original server keeps a sort of a reference to the
lower-level servers

• Root servers
– zone is a whole tree
– 13 in the world
• Primary server
– loads the information about the the zone from the
disk
• Secondary server
– loads the info from the primary server
• redundancy against failure

| U. K. Roy | [ 10/22/08 ]
 Web technology 39

Zones (cnt’d)

separate name
server
| U. K. Roy | [ 10/22/08 ]
 Web technology 40

DNS in the Internet

• Generic domains
– registered host according to their generic behavior
• Inverse domain
– used to map an address to a name
• Country domains
– the same format as in generic domain just 2 character
format
• us; nl; jp; fr; in
| U. K. Roy | [ 10/22/08 ]
 Web technology 41

Generic domain
Label Description

com Commercial organizations

edu Educational institutions

gov Government institutions

int International organizations
mil Military groups
net Network support centers
org Nonprofit organizations

aero Airlines and aerospace companies

biz Businesses or firms (similar to ‘com’)
coop Cooperative business organizations
info Information service providers
museu Museums and other nonprofit
m organizations
name Personal names (individuals)
pro Professional individual organizations

| U. K. Roy | [ 10/22/08 ]
 Web technology 42

Country domains

| U. K. Roy | [ 10/22/08 ]
 Web technology 43

Inverse domain
Example: a server wishes to
determine whether the client is
on the authorized list

• First-level node arpa for

historical reasons
• The servers are also
hierarchical
• Domain looks inverted
compared to a generic or
country domain

| U. K. Roy | [ 10/22/08 ]
 Web technology 44

Resolution
• Mapping a name to an address or vice-versa
• Resolver
– DNS client
– When a host needs to map an address to a name it calls
resolver that in turn access the nearest DNS server with a
mapping request
– A server either
• responds directly with an info, or
• refers the resolver to other servers
• asks other servers to provide info
• Recursive resolution
• Iterative resolution

| U. K. Roy | [ 10/22/08 ]
 Web technology 45

Recursive resolution
flits.cs.vu.nl -> linda.cs.yale.edu

• if the server is the authority for the domain name it checks its data
base and responds, otherwise
• it sends a request to another server…

| U. K. Roy | [ 10/22/08 ]
 Web technology 46

Iterative resolution
• The server returns either IP requested address or the IP
address of the server it thinks can resolve the querry

edu-derver.net

yale.edu

cs.vu.nl

| U. K. Roy | [ 10/22/08 ]
 Web technology 47

Dynamic DNS
• What if a new host joins the network or a host is
removed or an IP address is changed?
• DNS master file also has to be changed
• Changes so dynamic – a problem!

• Dynamic Domain Name System

• When a binding between IP address & host name is
determined (usually) DHCP informs DNS server

• Encapsulation
– DNS can use either UDP or TCP, using the well-known
port 53

| U. K. Roy | [ 10/22/08 ]
 Web technology 48

DNS Mesaages
Messages

Query Response

| U. K. Roy | [ 10/22/08 ]
 Web technology 49

Header Format

• Identification
– 16-bit field used by the client to match response with the query

| U. K. Roy | [ 10/22/08 ]
 Web technology 50

Flag Fields

• QR: Query/Response
• OpCode: 0 standard, 1 inverse, 2 server status
• AA: Authoritative
• TC: Truncated
• RD: Recursion Desired
• RA: Recursion Available
• rCode: Status of the error

| U. K. Roy | [ 10/22/08 ]
 Web technology 51

Resource Records
• Five tupple in the form
– Domain_name Time_to_live Type Class Value

Type Meaning Value

SOA Start of Authority Parameters for this Zone

A IP address of a host 32-bit Integer

MX Mail Exchange Priority, domain willing to accept

mail
NS Name Server Name of a Server for this domain
CNAME Canonical Name Domain Name
PTR Pointer Alias for an IP address
HINFO Host Description CPU and OS in ASCII
TXT Text Uninterrupted ASCII text

| U. K. Roy | [ 10/22/08 ]
 Web technology 52

Resource Records
$TTL 86400
@ IN SOA rose.itd.jusl.ac.in. rose.itd.jusl.ac.in. (
2006062101
3H
15M
1W
1D )

IN NS rose.itd.jusl.ac.in.
; IN NS galaxy.itd.jusl.ac.in.
IN MX 1 rose.itd.jusl.ac.in.
rose IN A 203.197.107.107
www IN CNAME rose.itd.jusl.ac.in.
mail IN CNAME rose
dns IN CNAME rose
gateway IN CNAME rose
hporacle IN A 172.16.6.97
lotus IN A 172.16.6.107
galaxy IN A 172.16.6.108
nfs IN CNAME lotus
dhcp IN CNAME lotus
nis IN CNAME lotus
| U. K. Roy | [ 10/22/08 ]
Simple Mail Transfer
Protocol(SMTP)
 Web technology 54

SMTP
• Provides electronic mail(email) services using email addresses
– Sending a single message to one or more recipients
– Sending messages that include text, graphics, voice and video
• Asynchronous service

| U. K. Roy | [ 10/22/08 ]
 Web technology 55

SMTP
• SMTP Client/Server
– User Agent(UA)
– Mail Transfer Agent(MTA)

| U. K. Roy | [ 10/22/08 ]
 Web technology 56

SMTP
• Relay MTA—used to store mail in an intermediate stage

| U. K. Roy | [ 10/22/08 ]
 Web technology 57

SMTP
• Mail Gateway—used when either side does not use
TCP/IP protocol

| U. K. Roy | [ 10/22/08 ]
 Web technology 58

SMTP(User Agent)
• Defined in SMTP without any implementation details
• Normally a program that provides an interface to send and receive mails
• Example
– Elm, Pine, MH, Berkley Mail, Zmail, Mush
– Eudora, Webmail etc.
• Sending Mail
– Envelop
– Message
• Header
• Body
• Receiving Mail
– UA checks mailbox periodically

| U. K. Roy | [ 10/22/08 ]
 Web technology 59

email format

| U. K. Roy | [ 10/22/08 ]
 Web technology 60

Addresses

| U. K. Roy | [ 10/22/08 ]
 Web technology 61

Delayed Delivery
• Sender-site Delay
– Sender site stipulates a spooling
system
– UA creates message and forwards it
to Spooling system to store
– MTA checks spool periodically for
new mail
– Delay depends upon following
conditions
• IP address of the server is obtained
through DNS
• Receiver is ready or not
– If the message can not be delivered,
it is returned to the sender
| U. K. Roy | [ 10/22/08 ]
 Web technology 62

Delayed Delivery
• Receiver-site Delay
– After receiving mail, it is
stored in respective mailbox
for reading
– Example
• Sendmail uses individual files
to store mails

• Intermediate Delay
– Mails can be stored by
intermediate MTAs to send
them when appropriate

| U. K. Roy | [ 10/22/08 ]
 Web technology 63

Aliases
• One-to-many Expansion
– Allows one name, called alias to represent several different
email addresses
– A list of email addresses is associated with the alias using a
database map
– If an alias is defined, mail destined to that name is sent to
every recipient’s of the list
– If not defined, mail is sent to the user only

| U. K. Roy | [ 10/22/08 ]
 Web technology 64

Aliases

| U. K. Roy | [ 10/22/08 ]
 Web technology 65

One-to-many expansion

| U. K. Roy | [ 10/22/08 ]
 Web technology 66

Aliases
• Many-to-one Expansion
– A user can have many different email addresses
– An alias database is used for this map
– Single mailbox is used
– Mails destined to all theses email addresses are sent to single user

| U. K. Roy | [ 10/22/08 ]
 Web technology 67

Many-to-one expansion

| U. K. Roy | [ 10/22/08 ]
 Web technology 68

Mail Transfer Agent(MTA)

• Actual mail transfer is done through MTAs
– Client MTA is required to send mail
– Server MTA is required to receive mail
– Example
• Sendmail, squirlmail etc.

| U. K. Roy | [ 10/22/08 ]
 Web technology 69

SMTP

| U. K. Roy | [ 10/22/08 ]
 Web technology 70

SMTP Messages
Messages

Commands Response

| U. K. Roy | [ 10/22/08 ]
 Web technology 71

SMTP Messages(Commands)
• Commands
– Commands are sent from
client to server
– First five are mandatory

| U. K. Roy | [ 10/22/08 ]
 Web technology 72

SMTP Messages(Responses)
• Responses
– Commands are sent from server to client
– 3 digit code of the following form
• 2yz(positive completion)
– Requested command has been successfully completed and new
commands can be started
• 3yz(positive intermediate response)
– Requested command has been accepted, but recipient needs more
information for completion
• 4yz(transient negative completion reply)
– Command has been rejected, but error is temporary. The command
can be sent again
• 5yz(permanent negative completion reply)
– Command has been rejected permanently. The command can not be
sent again during this session
| U. K. Roy | [ 10/22/08 ]
 Web technology 73

SMTP Messages(Responses)

| U. K. Roy | [ 10/22/08 ]
Figure 22-15 Web technology 74

Connection establishment

| U. K. Roy | [ 10/22/08 ]
 Web technology 75

SMTP
• An Example

| U. K. Roy | [ 10/22/08 ]
 Web technology 76

Connection Termination

| U. K. Roy | [ 10/22/08 ]
 Web technology 77

MIME
• SMTP uses NVT 7-bit ASCII character set
– Can not be used for languages that are not supported by 7-bit ASCII
characters. E.g French, German, Hebrew, Russian, Chinese,
Japanese etc.
– Can not be used to send binary data or audio or video

• MIME(Multipurpose Internet Mail Extension)

– A supplementary protocol that allows non-ASCII data to be sent
SMTP
– Can be thought of as software functions that transform non-ASCII to
ASCII and vice versa

| U. K. Roy | [ 10/22/08 ]
 Web technology 78

MIME

| U. K. Roy | [ 10/22/08 ]
 Web technology 79

MIME
• Defines five additional headers
– MIME-version
• MIME-Version: 1.1
– Content-Type
• Type of the data used in the body
• Content-Type: <type/subtype; parameters>
• Subtype
– Text, Message, Image, Video, Audio etc
– Content-Transfer-Encoding
• Encoding to be used
• Content-Transfer-Encoding: <type>
• Type
– 7bit, 8bit, binary, Base64 etc.
– Content-Id
– Content-Description
| U. K. Roy | [ 10/22/08 ]
Figure 22-20 Web technology 80

Base64

| U. K. Roy | [ 10/22/08 ]
Figure 22-21 Web technology 81

Quoted­printable

| U. K. Roy | [ 10/22/08 ]
 Web technology 82

Mail Delivery

| U. K. Roy | [ 10/22/08 ]
Code Division Multiple
Access(CDMA)
 Web technology 84

Agenda
BACKGROUND
THE CELLULAR SYSTEM
MULTIPLE ACCESS SYSTEMS
CDMA INTERNALS
FEATURES OF CDMA
ADVANTAGES OF CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 85

cdmaOne Subscriber Growth

History September 1997-March 2001

100,000,000
90,000,000
80,000,000
70,000,000
60,000,000
50,000,000
40,000,000
30,000,000
20,000,000
10,000,000
0
Se De Ma Ju Se De Ma Ju Se De Ma Ju Se De Ma
p- c- r- n- p- c- r- n- p- c- r- n- p- c- r-
97 97 98 98 98 98 99 99 99 99 00 00 00 00 01

Asia Pacific North America Caribbean & Latin America Europe, Middle East, & Africa

| U. K. Roy | [ 10/22/08 ]
 Web technology 86

CDMA: More Than a Decade of Success

The Voice and Packet Data Solution

89 90 93 94 95 96 97 98 99 00
Nov. 88 CDMA IS-95 Commercial Nov. 99 Over 74
CDMA standard CDMA launched in HDR million
Cellular issued So. Korea mobility subscribers*
Concept demonstrated
Korea selects PCS PrimeCo More than 90
CDMA turns up nation- 83 CDMA licensees
wide PCS service operators
in 14 cities in cdma2000 1x
35 nations* Commercial
Sprint PCS Commercial
selects HDR
systems in
CDMA for Standardized as
100 U.S. cities
nationwide 1x EV
Nov. 89
San Diego network Japan selects CDMA subscribers
CDMA CDMA reach over 12.5
Open U.S. PCS standard million in 30
Demonstration countries on
First commercial 5 continents*
-Power Control CDMA system
-Rake Receiver in Hong Kong using First HDR
-Soft Handoff QUALCOMM phones call demonstrated

| *U. K. Roy
According to |CDG, EMC [ 10/22/08 ]
 Web technology 87

Agenda
BACKGROUND
THE CELLULAR SYSTEM
MULTIPLE ACCESS SYSTEMS
CDMA INTERNALS
FEATURES OF CDMA
ADVANTAGES OF CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 88

Cellular Network

PSTN

MSC

Forward

Reverse

Cell Site

Cell Site

| U. K. Roy | [ 10/22/08 ]
 Web technology 89

Agenda
BACKGROUND
THE CELLULAR SYSTEM
MULTIPLE ACCESS SYSTEMS
CDMA INTERNALS
FEATURES OF CDMA
ADVANTAGES OF CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 90

CDMA
• ACCESS SCHEMES
– Frequency Division Multiple Access
(FDMA)
• Each station is allocated its own frequency.

– Time Division Multiple Access (TDMA)

• Each station is allocated its own time slot for transmission

– Code Division Multiple Access (CDMA)

• Use of encoding schemes to multiplex several signals.

| U. K. Roy | [ 10/22/08 ]
 Web technology 91

Frequency Division Multiple Access (FDMA)

H I I H HI

F re q u e n c y  C h a n n e l 1
GO
O G GO

F re q u e n c y   C h a n n e l  2
| U. K. Roy | [ 10/22/08 ]
 Web technology 92

Time Division Multiple Access(TDMA)

H I C o m m o n  F r e q u e n c y   C h a n n e l H I

u s e r 2 u s e r 1 u s e r 2 u s e r 1

I O H G

G O G O

| U. K. Roy | [ 10/22/08 ]
 Web technology 93

| U. K. Roy | [ 10/22/08 ]
 Web technology 94

Code Division Multiple Access(CDMA)

“Bonjour” “Selemat Datang”
“Hello” “Guten Tag” “Buenos Dias”

Common Frequency Channel

| U. K. Roy | [ 10/22/08 ]
 Web technology 95

CDMA
• CODING

| U. K. Roy | [ 10/22/08 ]
 Web technology 96

Agenda
BACKGROUND
THE CELLULAR SYSTEM
MULTIPLE ACCESS SYSTEMS
CDMA INTERNALS
FEATURES OF CDMA
ADVANTAGES OF CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 97

CDMA
• THE SPREAD SPECTRUM

– CDMA is a form of Direct Sequence

Spread Spectrum communications.
– three key elements:
• 1. The signal occupies a bandwidth much greater than necessary
– Benefits--immunity to interference, jamming and multi-user access
• 2. The bandwidth is spread by means of a code which is independent of
the data.
• 3. The receiver synchronizes to the code to recover the data.

| U. K. Roy | [ 10/22/08 ]
 Web technology 98

CDMA
• THE DIRECT SEQUENCE SPREAD SPECTRUM

– Example

| U. K. Roy | [ 10/22/08 ]
 Web technology 99

CDMA
• THE SPREADING PROCESS

| U. K. Roy | [ 10/22/08 ]
 Web technology 100

CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 101

Coding and Spreading

Tx 0  1  0  1  1  0  1  0 Rx
1  0  0  1  0  1  1  0

U s e r  1  D a t a :    0 1
B TS M o b ile U s e r   1  D a t a :  0 1
U s e r  2  D a t a :    1 0

U s e r   1  C o d e :   0 1 0 1 U s e r  1  C o d e :   0 1 0 1

U s e r   2  C o d e :   0 1 1 0

| U. K. Roy | [ 10/22/08 ]
 Web technology 102

Code Division Multiple Access

• Each station is assigned a sequence of numbers,
referred to as a “chip”.
– Examples:
A: +1, +1, +1, +1
B: +1, -1, +1, -1
C: +1, +1, -1, -1
D: +1, -1, -1, +1

– The chips’ sequences are carefully selected.

| U. K. Roy | [ 10/22/08 ]
 Web technology 103

CDMA
• The chip sequences are chosen to be pair wise orthogonal:
– Normalized inner product of any two chip sequences, S and
T(written as S.T) is 0. Mathematically

1 m
S.T = ∑ Si .Ti = 0
m i =m
• Following properties also hold

1 m 1 m 2 1 m
S.S = ∑ Si .Si = ∑ Si = ∑ (± 1) 2 = 1
m i =1 m i =1 m i =1
1 m 1 m
S.S = ∑ Si .S = − ∑1 = −1
m i =m m i=m
| U. K. Roy | [ 10/22/08 ]
 Web technology 104

Transmitting using CDMA

• Encoding rule for data stream:
– Data bit 0: encode as -1
– Data bit 1: encode as +1
– No data to send: encode as 0
• Transmission:
– Stations A, B, C, D each take their next data bit to send,
encode it as -1, +1, or 0; and multiply that code by each
number in the chip sequence to obtain a 4-tuple.
– The four 4-tuples are added together and the sum is
transmitted.
• The values will be the range -4 to +4, so 9 levels of physical
layer coding are needed.

| U. K. Roy | [ 10/22/08 ]
 Web technology 105

CDMA multiplexing

| U. K. Roy | [ 10/22/08 ]
 Web technology 106

Decoding CDMA
• The input to the demultiplexer is a 4-tuple of values
between -4 and +4.
• Each station takes the four values, and multiplies the
values by the chip sequence.
• The resulting values are then summed to obtain a single
value. The result will always be -4, +4, or 0.
• Divide the result by 4 to get a value -1, +1, or 0.
• Decode this result to a data bit of 0, 1, or no data.

| U. K. Roy | [ 10/22/08 ]
 Web technology 107

CDMA Demultiplexing

| U. K. Roy | [ 10/22/08 ]
 Web technology 108

CDMA
A: 0 0 0 1 1 0 1 1 A: (–1 –1 –1 +1 +1 –1 +1 +1)
B: 0 0 1 0 1 1 1 0 B: (–1 –1 +1 –1 +1 +1 +1 –1)
C: 0 1 0 1 1 1 0 0 C: (–1 +1 –1 +1 +1 +1 –1 –1)
D: 0 1 0 0 0 0 1 0 D: (–1 +1 –1 –1 –1 –1 +1 –1)

– – 1 – C S1 = (–1 +1 –1 +1 +1 +1 –1 –1)
– 1 1 – B+C S2 = (–2 0 0 0 +2 +2 0 –2)
1 0 – – A+B´ S3 = ( 0 0 –2 +2 0 –2 0 +2)
1 0 1 – A+B´+C S4 = (–1 +1 –3 +3 +1 –1 –1 –1)
1 1 1 1 A+B+C+D S5 = (–4 0 –2 0 +2 0 +2 +2)
1 1 0 1 A+B+C´+D S6 = (–2 –2 0 –2 0 –2 +4 0)
S1 C = (+1 +1 +1 +1 +1 +1 +1 +1)/8 = 1
S2 C = (+2 +0 +0 +0 +2 +2 +0 +2)/8 = 1
S3 C = (+0 +0 +2 +2 +0 –2 +0 –2)/8 = 0
S4 C = (+1 +1 +3 +3 +1 –1 +1 –1)/8 = 1
S5 C = (+4 +0 +2 +0 +2 +0 –2 +2)/8 = 1
S6 C = (+2 –2 +0 –2 +0 –2 –4 +0)/8 = –1
| U. K. Roy | [ 10/22/08 ]
 Web technology 109

CDMA
Proof:

S C = (A + B´ + C) C = A C + B´ C + C C = 0 + 0 + 1 = 1

• Generation of Orthogonal Chip sequences

– Walls Hadamard function

| U. K. Roy | [ 10/22/08 ]
 Web technology 110

CDMA System Block Diagram

A n ten na

V oic e A /D E n c o d e r  &
C o n v e rt er V o c od er
I n t e r le a v e r

D /A
S pre ad er
RF

Code
G en era to r

| U. K. Roy | [ 10/22/08 ]
 Web technology 111

| U. K. Roy | [ 10/22/08 ]
 Web technology 112

Vocoder (Voice Compression)

About 200 milliseconds

P u ls e   C o d e d
M o d u la tio n V o c od ed
(P C M ) V o ic e

  A /D VO CODE R 1  0  1  1  

| U. K. Roy | [ 10/22/08 ]
 Web technology 113

Digital to Analog Conversion

P u ls e   C o d e d
Vocoded D e m o d u la t i o n
V o ic e (P C M )

1  0  1  1 V OCO DER A /D

| U. K. Roy | [ 10/22/08 ]
 Web technology 114

Agenda
BACKGROUND
THE CELLULAR SYSTEM
MULTIPLE ACCESS SYSTEMS
CDMA INTERNALS
FEATURES OF CDMA
ADVANTAGES OF CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 115

Section Introduction

■ Universal Frequency Reuse
■ Power Control
■ Soft Handoff

| U. K. Roy | [ 10/22/08 ]
 Web technology 116

Frequency Planning Requirement

C
A

C DMA­201. 1 4

| U. K. Roy | [ 10/22/08 ]
 Web technology 117

CDMA Frequency Reuse

Traditional CDMA Systems

Cellular Systems

E F D A A A
C B E F A A A A
G C B A A A
D A G C A A A A
F D A A A A
B E F D A A A A
C B E A A A
A G C B A A A A
N=7 N=1

| U. K. Roy | [ 10/22/08 ]
 Web technology 118

Effective Power Control

Increased Power
Decreased
Power

Decreased Increased
Power Power

Decreased Power

Near/Far Problem
Increased Power Path Loss
Fading
| U. K. Roy | [ 10/22/08 ]
 Web technology 119

Effective Power Control—The Solution

• All users are controlled so that their signals reach the base station at
approximately the same level of power
• CDMA uses a 2-step Power Control process on the Reverse Link
– Estimate made by the mobile: Open Loop
– Correction supplied by the BS: Closed Loop
| U. K. Roy | [ 10/22/08 ]
 Web technology 120

Closed Loop Control

| U. K. Roy | [ 10/22/08 ]
 Web technology 121

Power Control During Soft Handoff

| U. K. Roy | [ 10/22/08 ]
 Web technology 122

Mobile Transmit Power Comparison

40

30
FM (AMPS)
20
Transmit
Power 10
CDMA
0

-10

-20 Average Transmit

Power
-30
0 50 100 150 200 250 300 350
CDMA 2 mW
Run Time in Seconds FM (AMPS) 700 mW

CDMA < 1/100 th

the power of FM
| U. K. Roy | [ 10/22/08 ]
 Web technology 123

Taking Advantage of Multipath

| U. K. Roy | [ 10/22/08 ]
 Web technology 124

Taking Advantage of Multipath

• MULTIPATH AND RAKE RECEIVERS

– Multipath signals are combined to

make a stronger signal
– Uses rake receivers—essentially
multiple receivers
• Each rake receiver gets different multipath signal and feeds them to
a central receiver to combine stronger multipath

| U. K. Roy | [ 10/22/08 ]
 Web technology 125

Multi Path Rake Receiver

0 Combiner
τ 1 Correlator 1
-10
τ 2 Correlator 2

dB -20 τ 3 Correlator 3

PN
Σ
Direct Multi
-30 XMTR
Path Paths

-40
τ ν Correlator n

-12 0 +12
Chips
1 Chip = 0.83 Microseconds

1. One of the receivers (fingers) constantly searches for different multipaths.

2. Each finger then demodulates the signal corresponding to a strong multipath.
3. The results are then combined together to make the signal stronger.

| U. K. Roy | [ 10/22/08 ]
 Web technology 126

Multi Path Rake Receiver

| U. K. Roy | [ 10/22/08 ]
 Web technology 127

What is Handoff?

Cell Cell
| U. K. Roy | [ 10/22/08 ]
 Web technology 128

The Need for Handoff

| U. K. Roy | [ 10/22/08 ]
 Handoffs in Analog and
Web technology 129

TDMA Networks

| U. K. Roy | [ 10/22/08 ]
 Web technology 130

Types of CDMA Handoff

• HANDOVER

– Hard Handover
• Break before make

– Soft Handover
• Make before break—possible a mobile station can be connected to
more than one BTS simultaneously
• Requires less power—reduces interference

| U. K. Roy | [ 10/22/08 ]
 Web technology 131

Types of CDMA Handoff

PSTN

BSC

BTS

BTS

PSTN

MTSO BSC

| U. K. Roy | [ 10/22/08 ]
 Web technology 132

Soft Handoff

Cell B
Cell A

Cell B
Cell A

Cell B
Cell A

| U. K. Roy | [ 10/22/08 ]
 Web technology 133

Soft Handoff
Frame Selection

BSC
Selector

| U. K. Roy | [ 10/22/08 ]
 Web technology 134

Soft Handoff Feature

• Made practical by frequency reuse = 1

• Process begun by mobile signal strength reports

• Determined by relative strength rather than absolute

threshold

• Two or more cell sites transmit to mobile

Mobile uses rake receiver to perform coherent
combining

| U. K. Roy | [ 10/22/08 ]
 Web technology 135

Hard Handoff vs. Soft Handoff

Continuity of call quality is maintained and
Dropped calls are minimized

Hard Handoff Soft Handoff

Analog, TDMA and GSM CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 136

CDMA Hard Handoff

FDMA
(Analog) CDMA

MTSO BSC

| U. K. Roy | [ 10/22/08 ]
 Web technology 137

Idle Handoff

Access
Pilot
BTS BTS

Pilot
Paging

| U. K. Roy | [ 10/22/08 ]
 Web technology 138

Agenda
BACKGROUND
THE CELLULAR SYSTEM
MULTIPLE ACCESS SYSTEMS
CDMA INTERNALS
FEATURES OF CDMA
ADVANTAGES OF CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 139

The 6 C’s of CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 140

CDMA Network Cost Advantage

Average Network Cost per Voice Minute of Use
$USD $0.078
$0.08 $0.065
$0.07 $0.056
$0.06 $0.047
$0.05 $0.039
$0.04
$0.03
$0.02
$0.01
$0.00 WCDMA TDMA GSM cdmaOne CDMA2000

Network Cost per Megabyte of Packet Data

Source: QUALCOMM Economic Model and White Paper,
$USD $0.47 “The Economics of Mobile Wireless Data,” February 2001,
<www.qualcomm.com/main/whitepapers/WirelessMobileData
$0.50 .pdf>
$0.45
$0.40
$0.35
$0.30
$0.25
$0.20
$0.15 $0.07 $0.06
$0.10 $0.02
$0.05
$0.00
GPRS WCDMA CDMA2000 1xEV-DO
| U. K. Roy | 1X [ 10/22/08 ]
 Web technology 141

The 6 C’s of CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 142

Voice Clarity Comparison

4
CDMA
7
GSM 13 kbps
RPE-LTP
CDMA 3
EVRC
GSM 5 CDMA
BER % 2 FER %

2.0 2.33 2.66 3.0 3.33 3.66 4.0

MOS
Voice clarity or speech quality is measured by
a Mean Order Score (MOS) and Bit/Frame Error Rates

| U. K. Roy | [ 10/22/08 ]
 Web technology 143

The 6 C’s of CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 144

Capacity is a CDMA Hallmark

AMPS = 1 GSM/TDMA CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 145

CDMA Capacity
FDMA

TDMA

CDMA

C = W log2(1 + S/N)

C = Capacity (bps)
W = Bandwidth (Hz)
S = Signal Power
N = Noise Power
| U. K. Roy | [ 10/22/08 ]
 Web technology 146

Frequency Reuse

n=7 n=1

CDMA Eliminates Frequency Planning

Higher System Capacity
Fast Deployment
| U. K. Roy | [ 10/22/08 ]
 Web technology 147

Soft Handoff Increases Capacity

Hard Handoff typically occurs farther away

from the serving
Base Station = More power required

 Lower system capacity

CDMA Soft Handoff typically begins

closer to the previous Base Station
which results in less power TX
 Higher system capacity
| U. K. Roy | [ 10/22/08 ]
 Web technology 148

Capacity Comparison

Sector Carriers/ Users/

Frequency Carrier Users/ Sector/ Sector/(Cell)
Technology Reuse Spacing Carrier 5 MHz 5 MHz

AMPS 7/21 30 kHz 1 8 8 (24)

GSM 3/9 200 kHz 8 2 16 (48)

TDMA (U.S.) 7/21 30 kHz 3 8 24 (72)
CDMA-Cellular 1 1.25 MHz 22 3 66 (198)
CDMA-1X 1 1.25 MHz 35 3 105 (315)
WCDMA 1 5 MHz 62 1 62 (186)

| Note:
U. K.ForRoy | the same frequency is reused in all 3 sectors.
CDMA [ 10/22/08 ]
 Web technology 149

The 6 C’s of CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 150

Coverage
• Link budget equal to or better than AMPS & GSM

• Due to:
 Spread-spectrum processing gain
 Strong channel coding, reducing Eb/No requirement

• Soft Handoff provides additional coverage gain

 Improved FER - e.g. 10% FER from each of 2 sites, combined
gives 1% FER.
 Theoretical 4.1 dB additional coverage
 In practice, up to 10dB coverage improvement in a fading
channel, depending on standard deviation of shadowing

| U. K. Roy | [ 10/22/08 ]
 Web technology 151

CDMA Link Budget

Link Budgets = Capital Budgets

dB dB
dB =
dB dB

• Link Budget advantage means

Bigger cell radius and greater

capacity per cell
• Fewer cells, fewer backhaul
• Less infrastructure to buy
• Faster time to market — fewer sites/permits needed
| U. K. Roy | [ 10/22/08 ]
 Web technology 152

CDMA Coverage

M
GS

CDMA
13 kbps
CD kbp
8
M s
A

Nominal cell radius

(900 MHz with 45 meter cell height)
Link Radius
GSM 143 3.6 km
CDMA (13 kbps) 148 5.1 km
CDMA (8 kbps) 150 5.9 km

| U. K. Roy | [ 10/22/08 ]
 Web technology 153

The 6 C’s of CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 154

Compatibility
1.25 MHz

CDMA2000
1xEV
3 Operators

7 Operators 5 MHz / 1.25 MHz

up to 2.4 Mbps

98 Operators
DATA
1.25 MHz 1.25 MHz
up to 2.0 Mbps
DATA
1.25 MHz DATA Up to 307 kbps
DATA Up to 64 kbps
14.4 kbps VOICE
CDMA is VOICE VOICE
INCREASED VOICE
CAPACITY
the Only 95A 95B CDMA2000 WCDMA /
Technology cdmaOne 1X / 1xEV CDMA2000

That Protects
Your Investment
A A B A B 1X/ A B 1X/ World
1xEV 1xEV Phone

1995 1999 2000 2001 2002 2003

In-band evolution within existing spectrum
| U. K. Roy | [ 10/22/08 ]
 Web technology 155

Compatibility
CDMA Mobile Terminals are Forward & Backward Compatible

cdmaOne Handsets Pin Compatibility: 3G Handsets

IS-95A to 1X

IS-95A/B to 1X

IS-95A/B to 1X

1x to 1xEV-DO

RF Compatibility: First commercial

No changes required for
cdma2000 1x
RF Front-end
handsets
Over 65 manufacturers
available now
...Just like the PC Industry
| U. K. Roy | [ 10/22/08 ]
 Web technology 156

The 6 C’s of CDMA

| U. K. Roy | [ 10/22/08 ]
 Web technology 157

Customer Satisfaction

■ Voice Quality

■ Battery Life

■ No Cross­talk

■ Privacy

| U. K. Roy | [ 10/22/08 ]
Switching Networks
 Web technology 159

Switching Networks
• Why switching networks?
– Two stations can communicate if they are connected
– Stations can be connected in two ways
– Directly
– Via switched networks

– Problems of connecting
stations directly
 Devices may be far apart—
expensive to setup a dedicated link
 A station may not require a link to
every other stations all the time
 No of links required is N(N-1)/2
 Cost grows with the square of the
number of devices

| U. K. Roy | [ 10/22/08 ]
 Web technology 160

Switching Networks
• Solution
– Attach a communication network—called switched network

– Stations—Devices need to communicate

– Each station is connected to a network node
– Network nodes forms the communication
boundary
– Purpose—to move the data from source to
destination
– Network can control the cost and connectivity

| U. K. Roy | [ 10/22/08 ]
 Web technology 161

Switching Networks
• Example

• Observations
– Some nodes connect only to
other nodes
– The sole task is the internal
switching of data
– Network is not fully connected
in general

– Node to node links are multiplexed links using either FDM or TDM

| U. K. Roy | [ 10/22/08 ]
 Web technology 162

Switching Networks
– Switching network
Circuit Switching
 Telephone network
Packet switching

| U. K. Roy | [ 10/22/08 ]
 Web technology 163

Circuit Switching Networks

• Characteristics
• Implies dedicated path between two stations
• Path is a connected sequence of links between network
nodes
• On each physical link, a channel is dedicated

• Communication involves three phases

– Circuit establishment
– Data transfer
– Circuit disconnect

| U. K. Roy | [ 10/22/08 ]
 Web technology 164

Single-Node Network
• Collection of stations are attached to a central switching node
• Central switch establishes a dedicated connection between two
devices that wish to communicate
• Digital switch
• Provides a transparent signal
path between any pair of
connected devices
• Network interface
• Hardware needed to connect
devices to the network
• Control Unit
• Establishes connection
(generally on demand basis)
• Maintains connection during
data communication
• Tears down connection
| U. K. Roy | [ 10/22/08 ]
 Web technology 165

Switch
• The switch hierarchy
– Five classes of
switching offices
• 10 regional offices
• 67 sectional offices
• 230 primary offices
• 1300 toll offices
• 19,000 end offices
– Calls are generally
connected at lowest
possible lable

| U. K. Roy | [ 10/22/08 ]
 Web technology 166

Digital Switches
• Digital switch
• Space Division switch
• Time division switch

• Space Division Switch

• Signal paths are separated
physically
• Crossbar switch
• Multistage switch

• Crossbar switch
– N input lines, N output lines
– N2 number of cross points
– Semiconductor switch is used to enable a cross point to connect an input to
output

| U. K. Roy | [ 10/22/08 ]
 Web technology 167

Limitations of Crossbar Switches

• Problems
– Number of cross points?
 Number of cross points grows with the square of the number of attached stations
– Only one path exists between pair of stations—Loss of cross points ?
 Prevents connection between two devices whose line intersect at that cross point
– Number of cross points used?
 Cross points are inefficiently used (at most N out of N2)

| U. K. Roy | [ 10/22/08 ]
 Web technology 168

Multi-stage Space Division Switch

• N input lines are broken into N/n groups of n lines
• Each group of n lines goes into a first stage switch
• Output of first stage becomes inputs to a group of second stage switch, and so
on
• Example

| U. K. Roy | [ 10/22/08 ]
 Web technology 169

Advantage of Multi-stage switch

• The number of cross points is reduced—increases crossbar utilization
– There are more than one path through the network to connect two
endpoints—increases reliability
– Number of cross
points?
– 1st stage: kN
2
N N N
– 2 stage:
nd k ×   ×   = k 
n n n

– 3rd stage: kN
2
N 
– Total: 2kN + k  
n
– Implication of K
– No of distinct paths from input to output
| U. K. Roy | [ 10/22/08 ]
 Web technology 170

Blocking & Non-blocking Switch

• Non-Blocking switch
• A path is always available to connect an input to an output
• Example—crossbar switch

– Blocking switch
– If one or more input-output pair can not be connected even if they are
available

– Example
– N=9, n=3, k=3
– Heavier lines indicate the
lines already in use
– Input line 9 can not be
connected to either 4 or 6

| U. K. Roy | [ 10/22/08 ]
 Web technology 171

Non-blocking Switch
• Condition for a switch to be non-blocking
• For a switch to be non-blocking
• k = 2n-1
• Total number of cross points in a three
stage switch 2
N
N x = 2kN + k  
n 2
N
N x = 2(2n − 1) N + (2n − 1) 
n
• Nx depends on number of switches (N/n)
• For optimal number of crosspoints
dN x
=0
dn
1
N  2
⇒ n =   and
2
( N x ) opt = 4N ( )
2 N −1
| U. K. Roy | [ 10/22/08 ]
 Web technology 172

Time Division Switch

• Time division switch
• TDM Bus switching
• Time Slot Interchange(TSI)
• Time Multiplexed Switch (TMS)

• Time Division Switch

• TDM concept
• N input and N output lines are connected
through controlled gates to a high speed
digital bus
• During a time slot input-output line gates are
enabled

• Number of cross
points? 4 N ( 2 N − 1)
• 2N instead of

| U. K. Roy | [ 10/22/08 ]
 Web technology 173

Time Division Switch

• Operation of TDM Bus switch
• 6 stations, 5 µs each
• Assume propagation time is zero
• 30 µs frame
• Control memory
• Indicates gates to be enabled
during successive time slots
• 6 words are needed

• Example
• During 1st time slot input gate
of 1 and output gate of 3 are
enabled

| U. K. Roy | [ 10/22/08 ]
 Web technology 174

Time Division Switch

• Number of cross points?
• 2N instead of 4 N ( 2 N − 1)

• Capacity?
• For 100 lines with 19.2 Kbps each, bus must be at least 1.92 Mbps

• Statistical TDM
• No fixed time slot for input, they are allocated on demand
• May be blocking
• Example:
• 200 stations 19.2 Kbps each
• Bus speed 2 Mbps
• About a half of devices can be connected at any time
• Varying data rate
• 9600-bps line gets one time slot while 19.2 Kbps line gets two time slots
• Circuit switching?
• TDM ?
| U. K. Roy | [ 10/22/08 ]
 Web technology 175

Time Division Switch

• Time Slot Interchange (TSI) Switch
• Operates by interchanging pairs of
slots
• n input lines, n output lines
• n input lines are scanned
sequentially to form an input frame
of n slots
• Slots are then reordered using a
time slot interchanger to make a
connection

• Example:
• Station 4 is connected to 0
• Station 7 is connected to 1

| U. K. Roy | [ 10/22/08 ]
 Web technology 176

TSI Mechanism
• Disadvantage
• Before constructing the output
frame, entire input frame mustr be
read—delay

• Example:
• n lines
• Memory access time is T µs
• Then time needed to process a
frame is 2nT
• For a frame period of 125 µs
and T=100 nsec
• number of lines that can be
allocated is 625

| U. K. Roy | [ 10/22/08 ]
 Web technology 177

TSI Operation with variable-rate input

• The number of slots to be used is stored in
channel assignment store
• Selector device at input uses no of time slots specified by
channel assignment store
• Input lines may be
sampled unequally, i.e.
more samples can be
taken from an input than
others

| U. K. Roy | [ 10/22/08 ]
 Web technology 178

Time Multiplexed Switch

Disadvantages of TSI switch
• TSI switches TDM data.
• TSI is simple to implement
• Size of TSI switch is limited by memory access time

• Example:
• Telephone line
• Bandwidth 4KHz/line
• Data rate 8Kbps/line
• Memory access time 100 nsec
• Maximum number of lines that can be allocated is 625

• Delay increases as the size of TSI switch grows

| U. K. Roy | [ 10/22/08 ]
 Web technology 179

Time Multiplexed Switch

• Solution
• To connect channels on different
TDM stream, space division
multiplexing is needed
• This technique is called Time
Multiplexed Switching (TMS)
• Multiple stage switch can now be
built by concatenating TSI and TMS
stages.
• Two stage TS switch is blocking
• Channel1,1 Channel2,3
• Channel1,2 Channel4,3
• To avoid blocking three or more stages
are used
• TST
• STS
• TSTST
| U. K. Roy | [ 10/22/08 ]
 Web technology 180

Time Multiplexed Switch

• Example

| U. K. Roy | [ 10/22/08 ]
 Web technology 181

Integrated Services Digital Network(ISDN)

• Primary public circuit switch—telephone network
• Designed for analog voice transmission
• Inadequate for modern communication needs
• a fully digital, circuit-switched network was built—Narrowband ISDN
• Primary goal was to integrate voice and non-voice services

• ISDN services
• Voice services
• Instant call setup
• Telephones that displays caller’s telephone number, name, address while ringing
• Call forwarding
• Conference calls worldwide
• Non-voice services
• Remote electric meter reading
• On-line medical, burglar, smoke alarms that automatically call the hospital, police
or fire department and give their address to speed up response

| U. K. Roy | [ 10/22/08 ]
 Web technology 182

Integrated Services Digital Network(ISDN)

• ISDN Architecture

• ISDN Interface
• The ISDN bit pipe supports following channels
• A – 4-kHz analog telephone channel
• B – 64 Kbps digital PCM channel for voice or data
• C – 8 or 16 Kbps digital channel
• D – 16-Kbps digital channel for out-of-band signaling
• E – 16-Kbps digital channel for internal ISDN signaling
• H – 384, 1536 or 1920-Kbps digital channel
• The ISDN bit pipe supports following channels
• Basic rate: 2B+1D
• Primary rate: 23B + 1D(U.S. and Japan) or 30B + 1D
(Europe)
• Hybrid: 1A + 1C

| U. K. Roy | [ 10/22/08 ]
 Web technology 183

Integrated Services Digital Network(ISDN)

• Broad band ISDN and ATM
• Operates at 155 Mbps—satisfying even video on demand
• Based on ATM technology—uses packet switching (it can emulate circuit
switching)
• Space division and time division switch can not be used for packet switching
• Switches should run at much higher speed
• Transmission in ATM Networks
• Uses fixed size cell (53 bytes)
• No requirement that cells rigidly alternate—cells arrive randomly from
different sources
• Normally uses Optical Fibre cable, but up to 100 meters coaxial cable can be
used

| U. K. Roy | [ 10/22/08 ]
 Web technology 184

ATM Switch
• Some input lines and some output (normally equal) lines

• ATM switches are synchronous—one cell is

taken from each input (if present)
• Switches may be pipelined—may take several
cycles before an incoming cell appears on its
output line
• Cells arrive at 150 Mbbps360,000 cells/sec
one cell must be taken every 2.7µs from every
input

• Common goal of any ATM switch

• Switch all cells with as low discard rate as possible
• Never reorder the cells on a virtual circuit

| U. K. Roy | [ 10/22/08 ]
 Web technology 185

Input queueing in ATM switch

• Problem arises when cells arriving at two or more input lines want to go to the
same output line

• Solution
1. Provide a queue for each input
line—if two or more cells collide,
one is chosen (randomly or
cyclically) for delivery, rest are
held for next cycles

• Head of line blocking—when a cell has to held up, it blocks rest of the cells behind
it even they could otherwise be switched
• To avoid head of line blocking a recirculating path can be used to send the losing
cells back to the input side
• Care must be taken to avoid out of order delivery

| U. K. Roy | [ 10/22/08 ]
 Web technology 186

Output queueing in ATM switch

1. Use queue on the output side

• Takes less cycles to switch all cells

| U. K. Roy | [ 10/22/08 ]
 Web technology 187

Knockout Switch
• Uses multiple limited number of output queue
• Concentrator selects a fraction of total cells eliminating (knockout) the rest

| U. K. Roy | [ 10/22/08 ]
 Web technology 188

Batcher-Banyan Switch
• Basic element

2x2 switch Cross parallel lower broadcast upper broadcast

0
0 0 0 0 0 0
1
4 4 2 2 1 1
2
1 1 1 1 2 2
3
5 5 3 3 3 3

4
2 2 4 4 4 4
5
6 6 6 6 5 5
6
3 3 5 5 6 6
7
7 7 7 7 7 7

| U. K. Roy | [ 10/22/08 ]
 Web technology 189

Batcher-Banyan Switch
000001 01 000010 02 000100 04
010011 23 001011 13 001101 15
100101 45 100110 46 010110 26
110111 67 101111 57 011111 37
Stage 0 Stage 1 Stage 2
0
0 0 0 0 0 0
1
4 4 2 2 1 1
2
1 1 1 1 2 2
3
5 5 3 3 3 3

4
2 2 4 4 4 4
5
6 6 6 6 5 5
6
3 3 5 5 6 6
7
7 7 7 7 7 7

| U. K. Roy | [ 10/22/08 ]
 Web technology 190

Example

3 000101
4
001000
5
010110
6
7

| U. K. Roy | [ 10/22/08 ]
 Web technology 191

Collision in a Banyan Switch

| U. K. Roy | [ 10/22/08 ]
 Web technology 192

Batcher Switch

| U. K. Roy | [ 10/22/08 ]
 Web technology 193

Routing in Batcher-Banyan Switch

| U. K. Roy | [ 10/22/08 ]
 Web technology 194

Switch

| U. K. Roy | [ 10/22/08 ]
B RIDGE
S
 Web technology 196

Introduction
• Many organizations have multiple (possibly different type) LANs
• Bridges can be used to connect them
• Operates at the data link layer
• Examples where bridges are used
5.
• Multiple LANs come into existence due to the autonomy of their owners
• Later there is a need for interaction, so bridges are needed
•
• Organizations may be geographically separated by considerable distance
• Cheaper connect them using bridges

•
• LAN is divided into separate LANs
to accommodate load

| U. K. Roy | [ 10/22/08 ]
 Web technology 197

Examples(cont.)
1.
• Limitation on the maximum physical distance between two machines in some LANs
• e.g. 2.5 Km for IEEE 802.3
• Only option is to partition the LAN and install bridges between segments
2.
• Reliability increases
• Bridges can be inserted critical places to prevent bringing down entire system
• Unlike repeaters, bridges can be programmed to exercise some discretion about what
it should forward and what it should not
•
• Security reason
• By inserting various places and being careful not to forward sensitive traffic, it is
possible to isolate parts of the network so that its traffic cannot escape and fall into the
wrong hands

| U. K. Roy | [ 10/22/08 ]
 Web technology 198

Operation of a Bridge

| U. K. Roy | [ 10/22/08 ]
 Web technology 199

Bridge from 802.x to 802.y

General Problems
• IEEE 802.x LANs use different Frame format

• —need reformatting during copying

• —requires CPU time, new checksum calculation
• —Introduces possibility of undetected errors

| U. K. Roy | [ 10/22/08 ]
 Web technology 200

Bridge from 802.x to 802.y

• Different data rate
• Slower LAN can not get ride of the frames as fast as they come in from a
faster LAN
• —buffer under run/run out of memory problem
• E.g. 802.4 to 802.3—802.3 operates slower than 10 Mbps due to collision

• Timer problem
• —faster LAN starts timer after forwarding a message to a slower LAN and
waits for the acknowledgement
• —timer expires before the message is delivered
• —source just retransmits the entire message increasing the load

• Different Maximum frame length

• 1500 bytes for 802.3, 8191 bytes for 802.4 and unlimited for
802.5(actually bounded by token holding time)
• Splitting the frame is not feasible as upper layer assumes that frames
either arrive or they do not and there is no provision reassembling frames
| U. K. Roy | [ 10/22/08 ]
 Web technology 201

Bridge from 802.x to 802.y

| U. K. Roy | [ 10/22/08 ]
 Web technology 202

IEEE 802 Bridge

• Transparent Bridge
• Features
• LANs connected via single bridge

• Transparent to the user—plug and play—no change in

hardware/software, no downloading of routing tables or parameters
• Operation of existing LAN is not affected
| U. K. Roy | [ 10/22/08 ]
 Web technology 203

Operation of a Transparent Bridge

• Operates in promiscuous mode
• Accepts every frame from all the LAN to which it is
attached
• On receiving a frame, it decides destination station is
• in same LAN—discard the frame
• on different LAN— forward the frame
• not known—use flooding
• Decision is done by looking a table

• Each entry of the lookup table is of the form

<Destination address, LAN address>
• Populated from incoming frames by backward learning
| U. K. Roy | [ 10/22/08 ]
 Web technology 204

Transparent Bridge
• Challenges
• Topology change
• Station moves from one LAN to another
• Attach arrival time in each entry of the lookup table
• Update it with new one
• Station is unplugged
• Scan the lookup table periodically and drop all entries a few minutes old

• Increased reliability
• Problems
• Cycle for ever
• Solution
• Spanning tree bridges

| U. K. Roy | [ 10/22/08 ]
 Web technology 205

Spanning Tree Bridge

• Example

• Spanning Tree formation

• Select a root—use flooding
• Use some distributed algorithm to form a spanning tree
• Algorithm continues to run to detect topology changes and updates the
spanning tree

| U. K. Roy | [ 10/22/08 ]
 Web technology 206

Source Routing Bridges

• Advantage of Spanning Tree Bridges

• Easy to install
• Plug and play

• Disadvantage of Spanning Tree Bridges

• Do not make optimal use of bandwidth—uses a subset of the entire
topology—spanning tree

• Relative importance of these two factors lead to split within 802

committees
• CSMA/CD and token bus people chose transparent bridge
• The ring people preferred a separate scheme called Source Routing
• Implementation complexity is put on the end stations rather bridges

| U. K. Roy | [ 10/22/08 ]
 Web technology 207

Source Routing Bridges

• Assumption
• Sender of each frame knows whether or not the destination is on its own
LAN
• Every machine in the internetwork knows, or can find, the best path to
every other machine

• Sending a frame to a different LAN

• Source machine sets the high-order bit of the destination address
to 1, to mark it
• It includes exact path the frame will follow in the frame header.

• Construction of path
• Each LAN has a unique 12-bit number (LAN id)—used to identify each
LAN uniquely
• Each bridge has a 4-bit number(Bridge id)—used to identify each bridge
in the context of its LANs
• Two bridges far apart may both have same number, but two bridges
between the same two LANs must have different bridge number
| U. K. Roy | [ 10/22/08 ]
 Web technology 208

Source Routing Bridges

• Construction of path(contd.)
• A route is then a sequence of bridge, LAN, bridge, LAN,…number
• Example
• Route from A to D would be (L1, B1, L2, B2, L3)

• Function of bridges
• A source routing bridge is only interested in those frames with high-order bit
of the destination set to 1
• For each such frame, it scans the route included in the frame header looking
for the number of LAN on which the frame arrived
• If the LAN number is followed by its own bridge number(i.e. the bridge is on
the path), the bridge forwards the frame onto the LAN whose number follows
its bridge number in the path
• If the incoming LAN number is followed by the number of some other bridge,
it does not forward the frame
| U. K. Roy | [ 10/22/08 ]
 Web technology 209

Implementation
• Software:
• Bridge runs in promiscuous mode, copying all frames to its memory to see
they have the high-order destination bit set to 1. If so, frame is inspected;
otherwise not

• Hybrid:
• Bridge’s LAN interface inspects the high-order destination bit and only
accepts frames with the bit set.
• easy to build into hardware and greatly reduces the number of frames the
bridge must inspect

• Hardware:
• Bridge’s LAN interface not only inspects the high-order destination bit, but it
also scans the route to see if this bridge must do forwarding
• frames that must actually be forwarded are given to the bridge
• requires complex hardware but wastes no CPU cycles as irrelevant frames
are screened out

| U. K. Roy | [ 10/22/08 ]
 Web technology 210

Source Routing Bridges

• Discovering routes
• If a destination is unknown, source issues a broadcast frame called
ROUTE DISCOVERY frame asking where it is
• This frame eventually reaches at the destination.
• Destination issues a ROUTE REPLY frame
• When reply comes back, bridges record (if it is not already recorded) their
identity in it
• First hop bridge inserts, incoming LAN number, bridge number and outgoing
LAN number
• Other Bridges insert bridge number and out going LAN number
• Original source can then see the exact route taken and choose the best
route

| U. K. Roy | [ 10/22/08 ]
 Web technology 211

Source Routing Bridges

| U. K. Roy | [ 10/22/08 ]
 Web technology 212

Source Routing Bridges

• Problem of route discovery
• Results frame explosion

• Example
• No of frames at in LAN N is 3N-1
• N=13, no of frames is more than half a million—causing congestion
• Solution
• When an unknown frame arrives, it is flooded, but only along spanning tree
—total volume of frames is linear with the size of the network not exponential
• Improvement
• Once a host is discovered a route to a certain destination, it stores the route
in a cache, so that the discovery process will not have to be run next time for
this destination.
| U. K. Roy | [ 10/22/08 ]
 Web technology 213

Comparison of 802 Bridges

Issue Transparent Bridge Source Routing
Bridge
Orientation Connectionless Connection-Oriented

Transparency Fully Transparent Not Transparent

Configuration Automatic Manual

Routing Sub optimal Optimal

Locating Backward learning Discovery frames

Failures Handled by bridges Handled by hosts

Complexity In the bridges In the hosts

| U. K. Roy | [ 10/22/08 ]
 Web technology 214

END

| U. K. Roy | [ 10/22/08 ]
 Web technology 215

Example

| U. K. Roy | [ 10/22/08 ]
 Web technology 216

Example

| U. K. Roy | [ 10/22/08 ]
H IGH-
S PE ED
LAN S
 Web technology 218

High-Speed LANs
• Motivation
• 802 LANs and MAN are (generally) based on copper wire
• Work fine for short distance and low speed
• For longer distance and high speed, optical fiber must be used
• Advantage of optical fiber
• High bandwidth
• Not affected by electromagnetic interference from heavy machinery, power
surges, or lightning
• Impossible to wiretap without detection—Excellent security

• High-Speed LANs
• FDDI (Fiber Distributed Data Interface)—uses optical fiber
• Fast Ethernet—uses copper wire

| U. K. Roy | [ 10/22/08 ]
 Web technology 219

FDDI
• Features
• Topology
• Ring topology
• Data rate
• 100 Mbps
• Distance
• 200 km
• Capacity
• 1000 stations
• Error Rate
• 1 out of 2.5x1010
• Usage
• Can be used as any of the 802 LANs
• Can be used as a backbone to connect copper LANs

| U. K. Roy | [ 10/22/08 ]
 Web technology 220

FDDI
• Cabling
• Uses multimode fibers
Multimode fiber
• Uses LEDs instead of laser
• Due to lower cost Single mode fiber
• Does not harm human body (eye)
• FDDI cabling consists of two fibers one transmitting clockwise and another
transmitting anticlockwise

• If one breaks, other can be used as back up.

| U. K. Roy | [ 10/22/08 ]
 Web technology 221

FDDI
• Cabling
• If both breaks at a point, two rings can be joined into a single ring
• Each station contains relays that can be used to join two rings or bypass
the station in the event of station problem

• It defines two classes of stations

• Class A—connected to both rings—fault tolerant—costly
• Class B—connected to only one ring—cheaper
• In the physical layer, 4 out of 5 encoding is used
• Saves bandwidth(100Mbps Manchester encoding requires 200 mega baud)
• Loss of self clocking. To compensate this long preamble is used. Clocks are
required to be stable at least 0.005 percent—maximum frame size is 4500 bytes
| U. K. Roy | [ 10/22/08 ]
 Web technology 222

FDDI
• Frame Format

• MAC Protocol
• Similar to 802.5
• To transmit a frame, a station must capture token. Then it transmits a
frame and removes when it comes back
• Difference
• Mac layer in FDDI puts a new token as soon as it has finished transmitting
its frames
• This is necessary to increase performance as the length of the ring could be 200
km long
• FDDI permits synchronous frames for circuit-switched PCM or ISDN data
| U. K. Roy | [ 10/22/08 ]
 Web technology 223

Fast Ethernet
• FDDI is too complex, costly due to the use of optical fiber
• Solution?
• Keep 802.3 as it was, but make it faster
• Redo it totally and give it lots of new feature such as real-time traffic and
digitized voice
• IEEE chose the first one for the following reasons
• The need to be backward compatible with thousands of existing LANs
• The fear that a new protocol might have unforeseen problems
• The desire to get the job done before the technology changed
• 802.3u evolves—called fast Ethernet
• Supports a data rate of 100 Mbps
• Uses hubs/switches—vampire tap or BNC connectors are not allowed
• Cabling
Name Cable Max. Segment Advantage
100Base-T4 Twisted pair 100 m Uses category 3 UTP
100Base-TX Twisted pair 100 m Full duplex 100 Mbps
100Base-F Optical fiber 2000 m Full duplex at 100 Mbps; long run

| U. K. Roy | [ 10/22/08 ]
N ETW ORK
S EC URITY
 Web technology 225

Athentication Protocols
• Authentication is the technique by which a process verifies that its
communication partner is who it is supposed to be and not an imposter
• Authorization/Authentication
• Authentication Protocol Model
• An initiating user (or process/party), say, Alice wants to establish a secure
communication with a second user Bob.
• Example
• Bob is a banker and Alice is a customer
• Alice starts out by sending a message either to Bob or to a trusted Key
Distribution Center(KDC)
• Several other messages will be exchanged during the communication
• As these messages are being sent, a nasty intruder, say, Trudy may
intercept, modify, or replay them in order to trick Alice and Bob or just to
gum up the works
• Nevertheless, when the protocol has been completed, Alice is sure she is
talking to Bob and Bob is sure he is talking to Alice
• They will establish a secret session key to encrypt messages that will be
exchanged during communication
| U. K. Roy | [ 10/22/08 ]
 Web technology 226

Authentication Based on a Shared Secret Key

• Assumptions:
• Alice and Bob already share a secret key, KAB (A for Alice and B for Bob).
• This shared key might have been agreed upon in person or in any event
not on the insecure network
• A Challenge response protocol

• Notation used:
• A, B are identities of Alice and Bob respectively
• Ri’s are the challenges, subscript being the challenger
• Ki’s are keys, i indicates owner,
• Ks is the session key

| U. K. Roy | [ 10/22/08 ]
 Web technology 227

Authentication Based on a Shared Secret Key

• Protocol 1 A
1. Alice sends her identity, A, to Bob 2
RB
• Bob chooses a challenge, a large

Alice

Bob
3
random number, RB, and sends it back KAB(RB)
to “Alice”
4
• Alice then encrypts the message with RA
the key shared with Bob and sends the 5
cipher text, KAB(RB) back. KAB(RA)

• Alice picks a random number, RA, and

1
sends it to Bob. A, RA
• Bob responds with KAB(RA).

Alice

Bob
2 RB, KAB(RA)
• Above protocol works but it contains
extra messages
• These messages can be eliminated by
combining information as 3
KAB(RB)
Is it an improvement over the original one?
No, by using reflection attack, Trudy can defeat this protocol
| U. K. Roy | [ 10/22/08 ]
 Web technology 228

The Refection Attack

Trudy can break it if it is possible to open multiple sessions with the bob at once
The reflection attack is as follows:
1 A, RT
• It starts out with Trudy claiming she is
2 RB, KAB(RT)
Alice and sending RT

Trudy

Bob
• Bob responds as usual with his own 3 A, RB
challenge RB
4 RB2, KAB(RB)
• Now Trudy is stuck. What can she do?
She does not know KAB(RB). She can open 5 KAB(RA)
a second session with message 3
supplying RB taken from message 2 as her
challenge

• Bob encrypts it sends backs KAB (RB) in message 4

• Now Trudy has the missing information, so she can complete the first session and
abort the second one. Bob is know convinced that Trudy is Alice

| U. K. Roy | [ 10/22/08 ]
 Web technology 229

The Refection Attack

Three general rules that often help to develop authentication protocols are as
follows:
• Have the initiator prove who she is before the responder has to. (In the above case,
Bob gives valuable information before Trudy has to give any evidence who she is)
• Have initiator and responder use different keys for proof.(This means having two
shared keys KAB and K’AB
• Have the initiator and responder draw their challenges from different sets. For
example, initiator must use even number and the responder must use odd number

| U. K. Roy | [ 10/22/08 ]
 Web technology 230

Establishing a shared key

Shared Secret key based authentication protocols assumes the existing of Shared
Secret Key
How can it be established?

Diffie-Hellman key exchange

• Assumptions:
• Alice and Bob have to agree on two large prime numbers, n, and g, where
(n-1)/2 is also a prime number.
• These number may be public.
• Alice picks a large (say, 512-bit) number, x, and keeps it secret. Similarly,
Bob picks a large secret number, y.
1 n, g, gx mod n
• Alice initiates the key exchange
protocol by sending Bob a message 2
Alice gy mod n

Bob
containing (n, g, gx mod n)

• Bob responds by sending a message Alice computes Bob computes

containing (gy mod n) (gy mod n)x =gxy (gx mod n)y =gxy
mod n mod n

| U. K. Roy | [ 10/22/08 ]
 Web technology 231

Establishing a shared key

Example
• n = 7, g = 3,
• Alice picks x = 8 and Bob picks y = 10

1 7, 3, 38 mod 7
• Alice initiates the key exchange
protocol by sending Bob a message
containing (7, 3, 38 mod 7) 2
310 mod 7

Alice

Bob
• Bob responds by sending a message
containing (310 mod 7) Alice computes Bob computes
(310 mod 7)8 =380 (38 mod n)10 =380
mod 7= 2 mod 7 = 2

• Is Diffie-Hellman algorithm secure?

• No, Bucket Brigade attack can break this algorithm.
• Basic idea
• When Bob gets the first message, how does he know it is from Alice?
• Trudy can exploit this fact to deceive both Alice and Bob.

| U. K. Roy | [ 10/22/08 ]
 Web technology 232

The Bucket brigade attack

• Alice and Bob picks x and y respectively
• Alice sends message 1 intended for Bob. Trudy intercepts this message in the middle

• Trudy picks z, and sends message 2 to Bob, using correct g and n obtained from
message 1. She also sends message 3 back to Alice
• Later, Bob sends message 4 to Alice which Trudy again intercepts and keeps.
• Now everybody does the modular arithmetic. Alice computes secret key gxz mod n so
does TrudyAlice thinks she is talking to Bob, so she establishes a session key
(with Trudy). So does Bob.
• Both are under illusion that they have a secure channel to each other, but actually
not
Alice Trudy Bob
picks x picks z picks y
1 n, g, gx mod n 2 n, g, gz mod n
3
Trudy
gz mod n
Alice

Bob
4
gy mod n

| U. K. Roy | [ 10/22/08 ]
 Web technology 233

Authentication using Key Distribution Center

• Problems
• To talk to n people n, shared secret keys are necessary.
• Key management would become a real burden
• Solution
• Introduce a trusted Key Distribution Center(KDC)
• Each user has a single shared key with KDC
• Authentication and session management now goes through KDC

| U. K. Roy | [ 10/22/08 ]
 Web technology 234

Authentication using Key Distribution Center

• Alice picks a session key and tells the KDC that she wants to talk to Bob using Ks

• This message is encrypted with the secret key KA Alice shares(only) with KDC
• KDC decrypts this message to extract Bob’s identity and session key
• It then constructs a new message containing Alice’s identity and session key and
sends this message to Bob.
• This message is encrypted with the secret key KA Alice shares(only) with KDC.
• When Bob decrypts this message, he learns that Alice wants to talk to him and which
key she wants to use.
1 A, KA(B, KS) 2 KB(A, KS)
Alice

KDC

Bob
• Is this algorithm secure?
• Answer:No, replay attack can break this algorithm

| U. K. Roy | [ 10/22/08 ]
 Web technology 235

The Replay Attack

• Trudy can figure out some legitimate service she can perform for Alice, makes an
attractive offer and gets the job
• After doing the work, Trudy politely requests Alice to pay by bank transfer.
• Alice then establishes a session key with her banker Bob.
• She sends Bob a message containing money to transferred to Trudy’s account
• Meanwhile, Trudy is back and she copies both message 2 and the message follows
it.
• Later she replays both of them to Bob.
• Bob thinks that Alice might have hired Trudy again. Bob then transfers an equal
amount of money from Alice’s account to Trudy’s account

1 A, KA(B, KS) 2 KB(A, KS)

Alice

KDC

Bob
| U. K. Roy | [ 10/22/08 ]
 Web technology 236

Solution to the Replay Attack

• Include a timestamp in each message
• Problem
• Clocks are never synchronized. Trudy can replay the message during this
interval and get away with it
• Put a one time unique message number, calld nonce
• Problems
• nonces must be remembered for ever. Trudy can try a 5-year old message
• If a machine crashes, nonces are lost.
• Timestamps and nonces can be combined to limit how long nonces have to be
remembered

| U. K. Roy | [ 10/22/08 ]
 Web technology 237

Needham-Schroeder authentication protocol

1 RA, A, B

KDC
2 KA(RA, B, KS, KB(A, KS))
Alice

3 KB(A, KS), KS(RA2)

Bob
4 KS,(RA2-1), RB

5 KS(RB-1)

| U. K. Roy | [ 10/22/08 ]
 Web technology 238

Authentication using Public-Key Cryptography

• Alice starts by encrypting her identity and
a random number, RA using Bob’s public 1 EB(A, RA)
key, EB
2 EA(RA, RB, KS)

Bob
Alice
• When Bob receives this message, he has
no idea of whether it came from Alice or
3 KS(RB)
Trudy
• So he sends Alice back a message
containing Alice’s RA, his own random
number, RB, and a proposed session key,
KS

• When Alice gets this message, she decrypts it using her private key. She sees RA.
This message must have come from Bob since Trudy has no way of determining RA.
Furthermore, it must be fresh not a replay since she just sent it.
• Alice agrees the session key by sending message 3

• Bob sees RB encrypted with the session key he just generated, he knows Alice got
the message and verified RA

| U. K. Roy | [ 10/22/08 ]
 Web technology 239

Digital Signatures

| U. K. Roy | [ 10/22/08 ]