This lecture is based on Wu Chapter 6.A Wide Area Network (WAN) is a
system which is capable of distributing information as far as it needs
to go - across the city, state or planet. A LAN, by contrast, is normally
considered to span a building, a campus or at most a corporate business
center.
Analog Telephony. POTS was designed for audio transmission,
obviously. The original microphones were magnetic, as were the earphones.
If a diaphragm is used to move a magnet next to a coil of wire, a varying
voltage is induced in the coil. Speech will cause the diaphragm to vibrate,
and a dynamic microphone is produced. However, the process is not
efficient, and so not very much sound is produced. This is the form of
telephone that Alexander Graham Bell invented. It actually had pretty good
frequency characteristics, for short distances. Dynamic microphones are
inherently rather linear (they don't distort signals much.)
Query 15.1. Alexander Graham Bell was working on his first telephone, between two rooms on the third floor of a warehouse in Boston. His assistant was in another room. Bell spilled a battery containing sulphuric acid, all over his pants. It started smoking and burning holes in his pants. How likely do you think it is, that Bell really said "Mr. Watson, come here, I want you!" as is historically reported?However, there was no form of amplification at that time and so the signal was quite weak; a better microphone was needed. Good old Tom Edison developed the carbon microphone which is based on a diaphragm which squeezes a collection of carbon particles. The resistance of the carbon varies with pressure, and so a rather large current from a battery could be used, instead of the tiny induced current from a dynamic microphone. However, carbon microphones had a much more restricted frequency response than dynamic microphones.
Fundamentally a telephone circuit involves a single loop of wire which begins at a battery (they still use batteries to power the phone system today because they're reliable and produce noisless direct current), passes through the trunk cables and into your microphone, through your earphone and back to the office. There the signal is amplified and passed on to switching equipment (by transformer) before closing the loop back to the battery. Your phone contains special circuits to reduce the "side tone" or amount of your own voice you hear in your own earphone, or you'd sound louder than the other party.
In the 1920's, scientists started studying the frequency characteristics of speech and discovered that if you transmitted sound in the range of 300 to 4000 hz, the speech was very intelligible. This bandwidth became the basis for the development of analog means of multiplexing signals - putting multiple signals through a single wire. Later, in the 1960's it also contributed to the development of digital signal transmission systems.
Telephonic digital lines (that's what the T stands for) come in T-1 through T-4 flavors, with T-1 and T-3 being the most commonly mentioned. US standard T-1 is about 1.5 megabits per second, and T-3 is about 45 mbps.
PBX - Private Branch Exchange. UCF, and every good-sized business,
has its own internal telephone system. UCF for instance has several thousand
telephones - but doesn't actually purchase several thousand separate phone
lines from the local carrier. Since it is known that an individual uses
the phone less than 1% of the time, 100 outside lines can serve the needs
of 10,000 people (these are example numbers, not real statistics.) Now
that PBX are largely digital, instead of "100 outside lines", a WAN technology
is used to communicate with the outside world.
a) If the angle of incidence is less than the critical angle, the light will cross the boundary, emerging at a different angle than it went in.
b) If the angle is greater than or equal to the critical angle, the light will bounce back. The important thing about this bouncing process is that it is 100% efficient (if the surface is smooth). No light goes across the surface.
This is the basis for fiber optic technology. A signal is inserted into a fiber (typically from a laser or light emitting diode), and then travels down a fiber, hopping from one bounce to the next. After a few kilometers it needs to be amplified and re-shaped (see next section.)
Dispersion. If the fiber were infinitely thin, it couldn't transmit much light due to its zero cross section. But if its cross section is non-zero, light can take multiple paths. Some of the paths are slightly shorter than others, and so an initially perfect square wave will degrade with distance into a somewhat dispersed pulse. Of course the pulse is also losing energy along the way if the glass is imperfect, as all things are. So it is necessary to provide an amplifier which detects the signal, thresholds it (detects its value above a certain minimum) and re-transmits it, cleaned up.
Multimodal fiber. The light in a single mode fiber is of a single wavelength, and that wavelength obviously determines the minimum pulse size. Light travels 300 million meters per second (3 * 10^9 m) in a vacuum, and significantly slower in other media. A fiber might achieve 200 million meters/second. If 500 nm light (wavelength = 500*10^-9 m=5*10^-7 m), this tells us that there are (3*10^9)/(5*10^-7) = 0.66*10^16 waves per second, which is quite a bit. However, due to dispersion, we cannot actually achieve proportionately high bandwidth.
To get maximum use of a fiber, it is possible to transmit many different wavelengths at the same time. This is "multimode" fiber. Theoretically, one multimode fiber can transmit all the communications traffic (telephone, TV, internet, the works) of the entire USA. (But it would be stupid to do so, wouldn't it?)
Query 15.2: (Based on the lecture.) What is the opto-electronic problem?
DCE: Data Circuit-Terminating Equipment: telco's modem or packet switch, which delivers data from the Public Data network (PDN) to our DTE.
Standards for WAN Packet Switching
X.25. International Telecommunications Union (ITU, formerly CCITT) standard; the big deal until Frame relay, SMDS and ATM came along. X.25 is connection-oriented, like a phone call. Corresponds to OSI layers 1 through 3. Virtual circuits are permanent or switched (SVC.)
X.25 was developed for analog copper circuitry and is is largely obsolete now, due to the arrival of fiber optics. X.25's error checking is too heavyweight for the more reliable fiber medium. X.25 used a checksum-and-retransmit protocol, but frame relay doesn't.
Key Fact: Used to be the main WAN technology; replaced by Frame Relay and ATM.
Frame Relay is also connection-oriented. It is used for permanent virtual circuits, such as linking a school's LAN to the county's WAN.Optical fiber has two great advantages over electronic (copper or microwave) means: (a) bandwidth, and (b) noise immunity. Fibers don't leak much, and they aren't susceptible to stray magnetic fields like electronic signals are. Frame relay relies on higher protocol layers (the "passengers") to handle error correction. Frame relay uses a CRC checksum to detect bad data, but just throws it away if it's bad.
Key Fact: Connection-oriented. Frame relay directly competes with ATM, and ATM is expected to eat its lunch.
SMDS - Switched Multimegabit Data Service - connectionless, Bellcore (formerly Bell Labs) system. SMDS addresses are like standard phone numbers, but the route can be "full mesh" - different packets in a message may travel via different routes.
SMDS is related to ATM via its 53 byte (48 payload + 5 header) cells; SMDS can be efficiently transmitted over ATM's "switching fabric" (hardware and software.)
Key Facts: (1) Mid-range of speeds (1 to 34 mbps); (2) SMDS does NOT provide QoS support, and thus is not suitable for high quality video or teleconference apps.
ISDN - the Consumer's Digital Service (maybe!) From early 1960's, the interoffice communication of POTS data has actually been digital; only the final mile was analog. ISDN is intended to make it digital all the way. ISDN is a connection-oriented service; in fact ISDN circuits have phone numbers.
Network terminator (plays the role of modem) is the NT1 box; but it's not really a modem because signal is digital on both sides.
Basic rate interface is 2 64 kbps B channels, one 16 kbps D channel.The D channel is used for "out of band" messages - that is, those which are not part of the data flow itself, but which concern startup and shutdown, error reporting and retransmission requests. You have to have Signaling System 7 (SS7) to use D for out-of-band; otherwise it's mixed in with B channels, which therefore only have 56 kbps. In addition to the OOB functionality, the D channel provides the user with 9.6 kbps for user data.
This is about as much as ISDN can put through a formerly-POTS twisted pair. Higher bandwidths require better wire.
The demo we saw on videotape of a teleconference from Lockheed-Martin Orlando to Troy, Alabama was running over three ISDN's which comprised six B-channels.
Key Facts: The predominant final mile service for digital information over existing copper.
SONET - Synchronous Optical Network - is today's Interstate Highway. The basic idea is that every 125 microseconds, one 8-bit sample from lots of signals is sent. Since 1,000,000/125=8000, this means 8000 8-bit samples per second, or 64 kilobits/second. Here "lots" means 86*9 = 851 channels in an STS-1 link. With three times the speed, you can combine three STS-1 links into an STS-3.
For voice, you don't really need 64 kbps, and so multiple voice signals can be multiplexed into one of the channels. Higher bandwidth applications might be spread across several channels. These feed-streams are called "virtual tributaries". They are folded into the overall stream by ADM's = "Add/Drop modules". Think of ADM as an on-ramp or off-ramp on the expressway. SONET circuits are hard-wired; you add and subtract traffic through ADMs, but you don't reconfigure the net except by physically pulling fiber connectors off of a box.
Query 15.3: Explain how the dual rings in Figure 6.13 provide high levels of network survivability.
BISDN - Broadband ISDN. Future stuff. "The master plan for an advanced digital telecom infrastructure...." Uses ATM over SONET. BISDN provides virtual circuits (of course it does, because ATM does!) Each STS-3 frame carries 44 ATM cells.
At the end of the BISDN section, the tag line: "Classical IP is the network protocol used for this application." Sort of like saying "the queen of England wears cotton socks from K-mart." But it works, eh? IP is the English language of cyberspace. Guess it's time to learn more about IP and the Internet. In fact that's the next chapter's topic.
There's not a lot of material to query about, in this chapter. Let's do some queries about Chapters 11, 12 and 13 in Dodsworth:
Dodsworth Chapter 11: PLACEHOLDER: Landscape and narrative in Virtual Environments. Brenda Laurel, Rachel Strickland, Rob Tow.
Query 15.4: Laurel says "In VR, one is not done unto, but doing. What opportunities for "doing" were provided in Placeholder? What were the designer's rationales (reasons) for providing these specific opportunities or affordances?
Query 15.5: What is a "smart costume" in PLACEHOLDER? Could a guest see the smart costume they were wearing?
Dodsworth Chapter 12: Beyond Shoot Your Friends. Celia Pearce.
Query 15.6: Pearce draws an analogy between violent video games and crack cocaine, saying "the market wants crack, but that doesn't mean we should make more of it." Develop one argument in favor of, and one argument against, this analogy.
Query 15.7: There's a country & western song that contains the line: "Old men sit around and talk about the weather; old women sit around and talk about old men." Does this concept relate to anything Pearce says. Does it suggest why girls may be less interested in video games than boys? Does it suggest possible themes for video games targeted at girls?
Query 15.8: Does Pearce make any assertions about the kinds of activities that appeal to females which explain the success of the Barbie Fashion Design CD? How about assertions that seem to predict less success for this product?
Dodsworth Chapter 13: Applying Game Design to Virtual Environments. Stephen Clarke-Willson.
Query 15.9: Cite one instance from an actual video game of each of the key game design principles that Clarke-Willson lists:
- third person presentation
- discovery and exploration
- movement versus animation
- player control
- the use of maps
- the use of "weenies"
- closed environments
- constant positive and sporadic negative feedback
- complexity management and slow bullets
Query 15.10: How does the author propose to solve the three problems he poses:
- lack of depth perception
- management of player viewpoint
- navigation and targeting support
Back to the course index
Back to the course syllabus
Back to the previous lecture
Onward to the next lecture