CS 447: NETWORKS AND DATA COMMUNICATIONS Spring 2012 – TR 3:30-4:45 PM – EB 0012 Course Description Concepts of networks and data communications. Networking protocols and architecture; data encoding and transmission; network management; and distributed applications. Prerequisites CS 314 and CS 340, both with a minimum grade of C. Textbook
CS 447: NETWORKS AND DATA COMMUNICATIONS. Spring 2012 – TR 3:30-4:45 PM – EB 0012. Course Description Concepts of networks and data communications. Networking protocols and architecture; data encoding and transmission; network management; and distributed applications. Textbook. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
CS 447: NETWORKS AND DATA COMMUNICATIONS
Spring 2012 – TR 3:30-4:45 PM – EB 0012
Course DescriptionConcepts of networks and data
communications. Networking protocols and architecture; data encoding and transmission; network management;
There are two basic models for conducting networked applications.
Client-Server Systems Peer-to-Peer SystemsTasks are partitioned between the requesters of the service
(the clients) and the providers of the service (the
servers).
Computers pool their resources (memory,
processing power, bandwidth) without any centralized
coordination.
Centralization lends itself to easier maintenance,
upgrading, and security, as well as better ease-of-use.
P2P systems are impacted less by machine failures and
actually benefit from expansion due to increased
resources.
CLIENT-SERVER VS. PEER-TO-PEER APPLICATIONS
Page 6Chapter 1CS 447
Client-Server: E-Mail, Web Service,
Database Access, File Processing,
Facebook
Peer-to-Peer: Content
Delivery, File Sharing, Skype
ACCESS NETWORKS: DIAL-UP CONNECTION
Page 7Chapter 1CS 447
Dial-up Internet connections utilize modems (modulators/demodulators) to convert digital data into analog
transmissions across telephone lines, and then back to digital signals at the receiver.
• Requires no infrastructure except a telephone network• Uses handshaking protocol to
establish connection prior to data transfer• Low bandwidth (56 kbit/sec)
inhibits use for streaming, on-line gaming, etc.
NetZero combats the bandwidth problem by prefetching data, compressing text files, and
exploiting the user’s cache to prevent redownloading.
ACCESS NETWORKS: DIGITAL SUBSCRIBER LINE
Page 8Chapter 1CS 447
DSL uses standard telephone lines to provide large throughput downstream (to the customer) for Internet data and small throughput
upstream (from the customer) for voice.• Their lack of symmetry causes most
DSL connections to be called Asymmetric DSL (ADSL)• A higher frequency band (>25kHz) is
used for the subscriber line, while a lower frequency band (<5kHz) is still used for POTS (Plain Old Telephone Service: voice, fax, analog modems)
In May 2011, AT&T instituted a 150GB use cap for its DSL service, penalizing customers who consume “a
disproportionate amount of bandwidth”.
ACCESS NETWORKS: CABLE MODEM
Page 9Chapter 1CS 447
Extra bandwidth on cable television systems may also be exploited to provide Internet access.
• Television channels require 6MHz bands; downstream channels are allocated 6MHz as well (with upstream channelsallocated only 2 MHz)
• Such cable systems are usually Hybrid Fiber-Coaxial (HFC), especially if they provide on-demand service.
ACCESS NETWORKS: FIBER-TO-THE-HOME
Page 10Chapter 1CS 447
Instead of using legacy coax systems to support broadband Internet access, FTTH systems provide fiber optic cable connections to
individual residences.• Active Optical
Networks route signals directly to the intended customer, and buffer signals coming from the customer.
• Passive Optical Networks broadcast the same signal to multiple users, using encryption to prevent eavesdropping, while some multiple access protocol is used to deal with signals coming from customers.
ACCESS NETWORKS: ETHERNET
Page 11Chapter 1CS 447
Local area networks are typically used to connect end stations to an edge router, which, in turn, connects them to the Internet.
• Ethernet, developed by Xerox in the late 1970s, is the most common LAN technology
• Ethernet uses CSMA/CD:o Carrier Sense (CS) – Each end
station “listens” for incoming messages, and for conflicting traffic whenever it wants to transmit
o Multiple Access (MA) – A single bus is used to carry all messages
o Collision Detection (CD) – When a station determines that its outgoing message has collided with that of another station, both stations wait a random amount of time and then retransmit
ACCESS NETWORKS: WIFI
Page 12Chapter 1CS 447
Wireless Internet access occurs via wireless local area networks or via wide area cellular networks.
• In a WLAN, users transmit to and receive from an access point which is wired to the Internet.
• A single WAP can usually handle about 30 end systems within a 100-meter radius.
• Cellular systems process messages through base stations in cell towers, which forward the wired signals through switching centers to the Internet.
• Greater distances, more sources of interference, and larger numbers of users complicate this approach.
ACCESS NETWORKS: WIMAX
Page 13Chapter 1CS 447
WiMAX (Worldwide Interoperability for Microwave Access) is being developed as a wireless alternative to DSL and cable modems for
“last mile” broadband access.• While WiFi provides Internet access via
an intervening WLAN, WiMAX provides direct Internet access.
• WiMAX is sometimes called a 4G technology, but it fails to meet the 1Gbps requirement of 4G (although the planned WiMAX 2 does).
• Similarly, WiMAX’s European counterpart, LTE (long Term Expansion) does not currently meet 4G specifications, but its next iteration, LTE Advanced, should.
CIRCUIT SWITCHING
Page 14Chapter 1CS 447
Traditional telephone networks use switching technologies based on dedicated circuits between communicating end stations, with
potential wasted bandwidth, but with guaranteed capacity and consistent delays.
Frequency-Division Multiplexing: Better line utilization via
simultaneous sub-band channels
Time-Division Multiplexing: Better line utilization via
sequential channels based on time slices
PACKET SWITCHING
Page 15Chapter 1CS 447
To make better use of communication lines, most modern protocols have transmitting stations break messages down into manageable
“packets” that are forwarded through the network and reassembled at the destination stations.
Datagram Packet Switching (used in IP)
allows each packet to find its own path to the
destination
Virtual Circuit Packet Switching (used in ATM)
sets up a single route for all packets in the message to
follow
NETWORK DELAYS
Page 16Chapter 1CS 447
Without the dedicated lines associated with circuit switching, packet-switched communication experiences inherent delays.
Processing DelayRouters need time to process the packet headers(encryption, error-checking, address translation)
Queuing DelayPackets must await their turn to be transmitted(one packet at a time; limited buffer capacity)
Transmission DelayPushing the packet’s bits onto the link takes time(switches store packets before forwarding them)
Propagation DelayTime it takes to traverse the links to the destination(directly proportional to the end-to-end distance)
• Format data according to syntactical and semantic negotiations between source and destination; encrypt and decrypt messages.
Presentation Layer
• Establish, maintain, and discontinue dialogues between a source and a destination; synchronize data transfer via checkpointsSession Layer
• Detect and handle end-to-end transmission errors; alter transmission rate when too much congestion is encountered.Transport Layer
• Route messages from their source to their destination; take measures to reroute traffic when congestion is encountered.Network Layer
• Detect and handle transmission errors between consecutive network nodes; control access to the shared transmission medium.Data Link Layer
• Transmit bits across physical medium; determine frequencies and amplitudes to use when transmitting; specify role of connector pins.Physical Layer
TCP/IP
Page 24Chapter 1CS 447
The Defense Advanced Research Projects Agency developed the five-layer predecessor to OSI in the
1970s.• Higher layer protocols such as FTP, SMTP, and HTTP, as well as libraries to provide the equivalent of OSI’s Presentation and Session functionality.
Application Protocols
• TCP (and its less reliable, connectionless counterpart, UDP) is responsible for providing end-to-end message transfer capabilities.
Transmission Control Protocol
• IP deals with two major tasks, host addressing and identification, and routing packets to their destination
Internet Protocol
• Also known as the Network Access Layer, these protocols specify the networking scope of the local network to which a host is connected.
Link Protocols
• TCP/IP does not provide hardware specifications, instead assuming that it can be implemented in a hardware-independent manner.
Physical Protocols
SECURITY ISSUES: MALWARE
Page 25Chapter 1CS 447
Various types of malicious software has been designed to disrupt operations or gather private information.
Trojan Horse; 69.99
Computer Virus; 16.32
ComputerWorm,7.77
Adware,2.27 Backdoor,
1.89
Spyware,0.08
Types of Malware
Trojan HorseDisguised as innocuous software, but concealing a
harmful payload when downloadedComputer Virus
A program that copies itself, infects a system, and then infects other systems via network access
Computer WormSelf-replicating software that spreads on a network without human intervention or attaching to other
softwareAdware
Software that automatically downloads and displays advertisements
BackdoorA method for bypassing normal authentication to
remotely access a computer
SpywareUndetected software that collects information about
users without their knowledge
SECURITY ISSUES:DENIAL OF SERVICE ATTACKS
Page 26Chapter 1CS 447
A DoS attack is designed to make a computer resource unavailable to its intended users,
typically by saturating the resources with
external communication
requests.
• The attacker contacts compromised “handler” machines.
• The handlers then contact pre-infected “zombie” agents..
• The zombies proceed to bombard the target server..
SECURITY ISSUES:PACKET SNIFFING
Page 27Chapter 1CS 447
Packet analyzer software is used to intercept and log traffic passing through
a network.Originally developed as a
means for network administrators to monitor system usage, diagnose
problems, and detect intrusion or misuse, this
software can itself be misused to spy on other
network users and to collect sensitive
information such as passwords or e-mail
messages.
SECURITY ISSUES: SPOOFING
Page 28Chapter 1CS 447
Because many TCP/IP protocols have no mechanism for authenticating source or destination information, it is possible for
someone to falsify data and masquerade as someone else, potentially accessing data illegitimately.
To determine how to send a message to destination B, source A sends a request
containing B’s IP address to a server, which broadcasts the request.
Hacker C responds to the server’s request, and the server naively associates B’s IP address with C’s location, so source A communicates with C,
thinking that it is B.
SECURITY ISSUES:MAN-IN-THE-MIDDLE ATTACKS
Page 29Chapter 1CS 447
An extension of the spoofing attack has the hacker pretending to the client to be the server, as well as pretending to the server to be the
client.
This not only permits the hacker to intercept messages between the client and the server, but also to inject false