Chapter 9, Communications and Networks

Discovering Computers 2004

Modified 11 May 04 1523 hrs.

Topics

Communication:
Process of exchanging a message between devices (including humans).
Anatomy of a communication system
Sending device: Message composition, coding
Communications device: transmitter, transducer
Communications channel: propagation in communication channel
Communications device: transducer, receiver
Receiving device: decoding, message display
Uses of communication
Email, voice mail, fax, smart phones, instant messaging, Web, electronic commerce, Internet telephony
telecommuting, audio/video data exchange, navigation, commerce
System design
System design tries to maximize service by
efficient use of transmission channel
transmission speed and number of messages on channel: channel bandwidth
effective use of transmission channel
keep signals detectable and recoverable
economic use of transmission channel
lowest cost methods consistent with operational constraints
System architecture is designed to meet the requirements for speed, reliability, security.
network topology
communication devices
communication protocols
network management software
Communication channel: transmission medium, energy path
conduction, radiation, constrained radiation
signal, interference (other signals), noise (random electromagnetic disturbance); 
Signal-to-Noise Ratio (SNR); Signal-to-Interference Ratio
Computer and Communication Technology Principles
Energy Category Technology or Device
Acoustic Acoustic Coupler Modem
Quartz Crystal Memory
Light Fiber Optics
Laser
Photosensitive Diode
CD-ROM
Infrared IrDA Port
Magnetic Magnetoelectronic Memory
Magnetic Bubble Memory
Magnetic Tape
Magnetic Disk
Electrical Conduction Solid Conductor Cable
Electromagnetic Radiation Radio
Microwave
Electronic Semiconductor Memory
Mechanical Impact Printer
Electrochemical Gas Plasma Display (Ions changing energy state)
(Neon signs)
Thermal Thermal Printer
 
Security and physics
Fiber optic cable is inherently more secure than solid conductor cable.
Electrical charge in motion generates a magnetic field.
Changing magnetic field can be detected by placing a coil in that field.
An electrical signal conducted through a cable can be detected by placing a coil external to the cable. 
Photons in motion do not generate a magnetic field.
Light propagating through an optical fiber cannot be detected by placing a coil external to the cable.  To intercept such a signal, you must cut the cable and insert a splitter.
Any time you cut a cable and insert a splitter (optical cable), or tap a cable (electrical cable), you change the impedance of the cable.  This change in impedance can be detected.
When a cable is broken, the location of the break can be approximately determined.
Sources of electromagnetic noise
Sunspots http://www.spaceweather.com/.  Sunspots discovered in 1613 by Galileo.
Sunspots are a major source of electromagnetic noise that interferes with radio communications. They also affect Earth's weather patterns.
April 2000 maxima sun picture and plot http://science.msfc.nasa.gov/ssl/pad/solar/images/ssn_predict_l.gif 
Plot of sunspot numbers since 1749 A.D. (27 years before the American Revolution... when Dr. Ben Franklin was professionally active.) http://www.sunspotcycle.com/images/zurich.gif 
Year 2000 A.D. is a solar maximum.
Electric motors, transformers
Electrical switches changing  position
Spark plugs in a car, the associated electrical pulses in spark plug wires (which act like a transmitting antenna), ignition coil
Household electrical appliances (can opener, microwave oven, vacuum cleaner, dish washer, clothes washer and dryer, air conditioner compressor, refrigerator compressor, ham radio transmitter)
Cables
Shielding is a conductor. 
Used to protect signal from noise external to a conductor.
Soft steel (not aluminum or copper) shielding also protects against external magnetic fields by giving magnetic flux an alternate path.
Sometimes a shield is used as a conductor also.
Insulation is a nonconductor, dielectric. 
Used to prevent electrical grounding and shorting, and for safety.
Armor: Used to physically protect a cable from 
debris (sparks, etc, in a factory)
impact
compression
burrowing creatures (like groundhogs: the major problem MCI has with buried cables in Texas)
shrapnel
Strength Member: Used to permit stringing cable for longer distances between supports by
supporting tension and 
reducing bending, to 
reduce or prevent stretching of the conductor or optical waveguide.
Moisture Barrier
prevent conducting fluids from permeating cable and causing shorts.
Conduction: 
conductivity or absorption coefficient, insulation, shielding
Twisted pair, shielded or unshielded
Twisting pairs of wires is done to reduce crosstalk.
Grounded shielding is used above 1 MHz to further reduce crosstalk, which is the dominant concern in system design above this frequency. 
Guided Waves (Constrained Radiation):
coaxial cable: 
Definition of coax cable: Multiple conductors symmetric about a common axis.
Some coax cables have air core, with cylindrical conductors separated by spacers.
Used for communication in the MHz band, individual conductors shielded to reduce noise, with insulating material separating conductor and shielding.
Fiber optic cable: light wave propagation
Free Waves (Unconstrained Radiation):
Radio Wave Propagation
Tutorial for Digital Designers: http://www.tapr.org/tapr/html/ve3jf.dcc97/ve3jf.dcc97.html 
Radio Propagation Conditions: http://www.hamradio-online.com/propagation.html 
VHF / UHF Tropospheric Ducting Forecast http://www.iprimus.ca/~hepburnw/tropo.html 
VHF - UHF Information http://www.rac.ca/6metrep.htm 
Ham Radio on the Net http://www.eham.net/links/results 
Physical principals: http://ecjones.org/physics.html  Dr. E. C. Jones home page http://ecjones.org/index.html 
Broadcast radio
Bluetooth: short range (10 - 100 m) wireless transmission
http://grouper.ieee.org/groups/802/15/ 
Specifications are available from http://www.bluetooth.org/specifications.htm 
Operates in the 2.4 GHz Industrial, Scientific, Medical band: 2.4 - 2.4835 GHz.
The standard is not world-wide.  France is incompatible with USA and the rest of Europe.
Channel spacing is 1 MHz.
3 classes of power: (1) 1 - 100 mW, 0.25 - 2.5 mW, and 1 mW.
The signal uses Gaussian Frequency Shift Keying modulation.
Bluetooth application specifications exist for the following:  generic access, service discovery application, cordless telephony, intercom, serial port, headset, dial up networking, fax, local area networking, generic object exchange, object push, file transfer, synchronization.  For details, consult http://www.bluetooth.org/docs/Bluetooth_V11_Profiles_22Feb01.pdf 
Cellular radio: Code Division Multiple Access (CDMA)
Microwave radio: 1 mm (300 GHz) - 30 cm (1 GHz) wave band [when frequency decreases, wavelength increases], 5 degree beamwidth, 12 W, line of sight transmission; relay at 30 mile intervals; atmospheric moisture is limiting factor
Infrared: 0.7 - 1000 mm wavelength, line of sight
Transmission links:
Communication satellite, geosynchronous orbit 
Satellite Terminology:  http://www.telesat.ca/satellites/term.html 
Antennae
Signal characteristics
Analog (page 3.11 and 9.12), digital (page 3.11 and 9.12)
In the real world, all signals are analog signals.
Digital signals
An ideal digital signal is a discrete signal. It can have possibly an infinite number of states. 
A real digital signal is an analog signal that is used to represent a discrete number of states.
HDTV 8-VSB is an example of a real digital signal that is not a binary signal.
Binary signals
An ideal binary signal is a digital signal signal that has only two states.  
A real binary signal is an analog signal that is used to represent two states.
Broadband is the opposite of narrowband.
The decision point of what constitutes broadband and narrowband depends upon the discipline or application.
An ideal narrowband signal is a pure sine wave.  Ideal narrowband signals do not exist in real life, but some are very, very close.
Real narrowband signals have some width Δf = fhi - flow, usually determined at the half-power points.  A signal is classified as narrowband depending upon the ratio  
fc / Δf.  Most people would agree that when this ratio is on the order of 0.001, the signal centered at frequency fc is a narrowband signal.  Generally, a signal that is a one-tenth decade or is a one-third octave band is not considered narrowband.
A signal is broadband if it is not narrowband.
Baseband is the opposite of a modulated signal.
A baseband signal is an unmodulated signal.
An AM radio signal is a modulated signal.  When you tune to 1500 kHz on your AM dial, the 1500 kHz is the carrier signal [c(t)].  The voice you hear [x(t)] is the intelligence-bearing signal.  The signal that is actually transmitted [y(t)] is the time product of these two signals.  This product is the modulated signal.
y(t) = x(t) * c(t)
x(t) is the baseband signal.  
The baseband signal is usually assumed to be the intelligence bearing signal.
y(t) is the modulated signal.
c(t) is the carrier wave.  
The carrier wave is usually assumed to be a much higher frequency than the baseband signal.
The carrier wave is chosen to match the signal to the propagation medium and for security.
By modulating the intelligence-bearing signal onto a carrier wave, you can transmit more intelligence-bearing signals through a channel simultaneously by using a different carrier wave for each intelligence-bearing signal.
There are other modulation schemes.  AM (amplitude modulation) and FM (frequency modulation) are the most well known examples to the public.  In digital communications, modulating the phase is an approach.  You can achieve parallel transmission by assigning a different frequency to each bit position.  Binary signals are modulated in a variety of ways, depending upon the application.
A more detailed illustration of Digital Signal Formats.
Transmission modes: 
asynchronous: start bit, stop bit
synchronous: time-synchronized transmitter and receiver 
Error detection and correction
Error detection without correction codes requires fewer bits than error detection with correction. When an error is detected, request retransmission.
parity bit: EVEN or ODD parity. An extra bit is added to the transmitted word to form an extended word.
EVEN parity: the extra bit is set to make the number of "one" bits in the extended word EVEN.
ODD parity: the extra bit is set to make the number of "one" bits in the extended word ODD.
Procedure:
Compute the parity bit, using the chosen convention (even, odd).
The message word and the originally computed parity bit are both transmitted.
The recipient recomputes a parity bit on the received message word.
The recipient compares the transmitted parity bit with the newly computed parity bit.
If the parity bits are different, it is guaranteed that the received message is different than the transmitted message.
If the parity bits are identical, if there is a mistake, then an even number of mistakes were made. The usual assumption is that no mistake is made. For short transmissions over reliable and controlled paths, this is usually a valid assumption.
The parity bit technique is often used between memory and the CPU.
The parity bit technique is a special case of the more general check digit approach.
check digit
Perform a computation on a string of numbers that are to be transmitted. The result is called a check digit.
Transmit the result of this computation along with the original string of numbers.
The receiver recomputes the check digit.
Compare the recomputed check digit to the transmitted check digit. If the results are different, it is guaranteed that an error occurred. The ability to learn more about the error depends on the method of computing the check digit and the data.  If the results are identical, an error is still possible, but is much less likely.
Hamming code: single-error-correcting code, a type of Linear Block Code
W. W. Peterson and E. J. Weldon, Jr., Error Correcting Codes, 2nd ed., Wiley, New York, 1972.
S. Lin, An Introduction to Error Correcting Codes, Prentice-Hall, Englewood Cliffs, N.J., 1970.
Transmission direction: simplex, half-duplex, full-duplex
Simplex is only one-way communication, like broadcast radio or TV.
Full-duplex is simultaneous two-way, like a telephone conversation.
Half-duplex is two-way, but only one-way at a time, like polite conversation, or transmission on a police radio or CB.
Transfer rates: bits per second (bps), bandwidth (Hz)
Nice tutorial: http://w3.sba.oakland.edu/netmgt/3nm_concepts/3nm_analog/sld001.htm 
Networking basics tutorial http://www.techtutorials.com/tutorials/networking.shtml 
Circuit-switching versus packet switching
Circuit-switched network maintains a connection for the duration of a message.
Synchronous communications between end-points is possible over a circuit-switched network.
Packet-switched network breaks messages into packets which are individually routed through the network and reassembled at the receiving end.
Synchronous communications between end-points is not possible over a packet-switched network.
Public switched telephone network (300 - 3300 Hz)
Circuit-switched network.
Designed for human voice communications.
Original infrastructure designed as analog system.
"The Last Mile Problem": much of the system from end-user to first connection is still not digital.
Dial-up line
Dedicated line, leased line
Integrated Services Digital Network (ISDN): Basic Rate Interface (BRI) for 3 signals, Primary Rate Interface (PRI) for 24 signals.
Digital Subscriber Line (DSL)
Asymmetric digital subscriber line (ASDL): fast receive, slow send. Surfing.
Faster receive: 1.54 Mbps-8.45 Mbps
Slower transmit: 128 kbps-640 kbps
xDSL: DSL with dial tone for voice communications.
T-Carrier line: multiplexed
T-1 line: 24 signals at 64 kbps each
T-3 line: 28 times the T-1 line capability (28 x 24 = 672 signals = 43 Mbps)
Line Costs: Charles Waltner, "Meet Your Connection", Fortune Technology Review, Technology Issue, pp 59 - 66 (Summer 2001).  Compare prices using $ / Mbps.
Line Cost Capacity Unit Cost
DSL private $ 40  / mo 5 - 20 times dial-up  
DSL business $ 100 - $ 300 / mo    
ISDN      
T1 $ 1200 / mo 1.54 Mbps $ 780 / Mbps
T3 $ 18,000 / mo 45 Mbps $ 418 / Mbps
Gigabit Ethernet
within 0.5 mi of optic fiber cable
$ 6000 / mo 10 Mbps $ 600 / Mbps
Optic: Time-Warner $ 400 / mo 400 kbps $ 1000 / Mbps
OC-3 $ 72,000 / mo 156 Mbps $ 462 / Mbps
Communication software
dialing, file transfer, terminal emulation, Internet access
Hyperterm, PC Pro
Communication devices and drivers
Modem: modulate, demodulate; 28.8 kbps, 56 kbps; internal, external; FAX; DSL
Consumer intro to modems: http://www.teleport.com/~curt/modems.html 
Cable modem: use with CATV; internal, external; 500 kbps, 2 Mbps; set-top box
Multiplexer, MUX: Time division, frequency division, phase shift
Network interface card (NIC), or network adapter: Installed in computer to interface it with the network. It includes the connectors for network cables.  Associated network adapter driver (software) must also be installed on the computer to communicate between the network adapter and the protocol driver (or transport protocol) (software) on the computer.
Choose a network adapter that is compatible with one of your computer buses.
Ethernet is low cost, simple, and satisfactory for 5 computers in a peer-to-peer topology.
Network hardware
Repeater: receive, amplify, retransmit
Hub (multiport repeater), concentrator, multistation access unit (MAU)
Bridge: connect LANs using same protocol. Can isolate local traffic on a LAN.
Switch: multiport bridge
Router: message switching device, chooses route for messages
Gateway: hardware and software to connect networks using different protocols
Network use
Hardware sharing
Data and information sharing
Software sharing
Communication
Network terminology
A network consists of a collection of nodes interconnected by links for the purpose of transmitting messages.
A message can be a single bit, a byte, a record, a file, or other data unit.  In a road network, a message can be a car or truck.
A link is a path, branch, or edge along which signals travel.  In a road network, a link is a road.
A node is a junction point of zero or more links.  In a road network, a node can be an intersection or destination point.
Network topology, network architecture
Network physical topology is the arrangement of cables and physical connections between devices in a communications network. 
Network logical topology describes the arrangement of which other processors each processor on a network shares information with for the purpose of processing.
Bus, ring, star, mesh
It is common to use a physical star topology to implement a bus, ring or star logical topology.
Communication protocol: set of rules and procedures for exchanging information among computers
Ethernet: Carrier Sense Multiple Access / Collision Detection (CSMA / CD); based on bus topology
Token ring: token passed around ring. Only device with token can transmit.
Packet switching: TCP/IP: Transmission Control Protocol / Internet Protocol
TCP/IP tutorial http://www.techtutorials.com/tutorials/tcpguide.shtml 
Wireless Application Protocol (WAP): for mobile devices.
Windows 98 includes the following network protocols: TCP/IP, IPX/SPX-compatible (Novell NetWare Internetwork Packet Exchange / Sequential Packet Exchange), NetBEUI (NetBIOS Extended User Interface: Windows for Workgroups, Windows NT Server, LAN Manager), 32-bit DLC (Data Link Control: IBM host, AS/400 over token-ring, FDDI, or Ethernet).
Network operating system
Administration
File management
Printer management
Security
Types of networks
Relationships
Peer-to-peer: share hardware resources and files by direct access
Client - Server network: file server, print server, database server, thin server (home use: internet access), telnet, ftp; WWW
More about different types of servers at http://serverwatch.internet.com/articles/servertruths/index.html 
Master - Slave network
Supercomputer architecture
Distributed processing
Internet, intranet, extranet
intranet: enterprise network; 
extranet
Area of Service
Home: share hardware, phone line
Local Area Network (LAN):  simple or sophisticated
Wide Area Network (WAN): interconnects LANs that are usually (but not necessarily) separated by large distances.
Metropolitan Area Network (MAN): WAN that serves a metropolitan area.
Firewall: hardware and software that acts as a security barrier between networks, permitting conditional access to authorized users of various categories, and denying access to all other users.

Questions

What different media can digital signals be transmitted?
How do messages travel across the network?
What resources can be shared on a network?
What are different types of local area networks?
What are different types of network topologies?