Welcome to this course in the technology of telecommunications. We expect readers at both ends of the science and technology pre-knowledge scale, and in between. At one end of this scale is the average non-technical consumer or subscriber to a telecom service such as voice telephone. Most consumer telecommunications equipment, such as a wired or wireless telephone handset, is designed so you don't need to know how the internal mechanism works in order to use it. At the other end of the scale is the engineer who needs to design all or part of a telecom system. Most readers of these notes are somewhere in between these two extremes. Because of your work or personal interest, you need to know more about the technology of telecommunication equipment; what are its capabilities and limitations? Another area we will touch on briefly in this first lecture only, because it diverges from the main topics of the courrse, is the followoing: How does an executive or a venture capital investor in a new business decide which projects should be funded? Do all the people involved in the decision know the correct technical facts about the project? Is the product price in the range that buyers are willing to pay? (We will discuss this last question in some detail.) You can develop a reputation for doing a good job even when the project you are working on ultimately is not successful, but your career will most likely advance much further if you are associated with a success.
2 2 Release. 1 ; Aug. ‘97 These notes include several historical summaries of technological advances in telecommunication, but our main emphasis is on the development of digital technology for multiplexing, switching and transmission. Most of this “digital history” has been made since 1961, when the first T-1 digital multiplexer was made public by Bell Laboratories and Western Electric, then respectively the R&D organization and manufacturing organization of the conglomerate telecommunications firm American Telephone & Telegraph (AT&T) ¹. The technology of this particular design was copied under license throughout North America and Japan. Similar digital multiplexers, called E-1, were developed in Europe soon after, and are used in PSTN networks in other parts of the world. The favorable economics of digital multiplexers caused rapid displacement of earlier analog multiplexers, and made the introduction of digital telephone switches during the 1970 and 1980 decades. Today virtually 100% of the telephone lines in developed countries operate with digital switching, multiplexing and transmission equipment. Digital telecom technology has produced tremendous improvements in telecommunications reliability, fidelity (accuracy of the signal information at the destination) and lower costs, compared to the prior analog transmission and electro-mechanical switching technology. Although the manufacturing sector of the telecom industry is in an economic depression as this is written (year 2006), telecom service and equipment are products that will still have a major and growing future market. ______ Note 1: The AT&T corporation formed in 2005 via a merger of several regional telecom service providers is a business descendant of the AT&T corporation mentioned above, but is not precisely the same firm.
First course assignment: Print on paper, fill in and send (via paper or e-mail) to the instructor the Student Information form (from www.engr.smu.edu/~levine/ee8320). Regarding term paper: To obtain approval for your term paper topic, send an abstract or outline or table of contents (TofC), via e-mail to the instructor, richard.levine@ gmail.com, not later than Oct. 22, 2006.????? These message can be read most rapidly if your outline/abstract is in the body of the e-mail, but file attachments are permissible. An abstract is approximately one page or less, summarizing your entire paper (including the conclusion – assuming that you know the conclusion when you write the abstract). You may send revised outline or abstract or TofC if the details of the paper change significantly as you research and write it. About 2/3 of all outlines are approved without change. About 1/3 are changed in various ways. For example, some topics are too new to have much solid factual information available, or the topic or treatment is not suitable for the educational and experience background of the student. The average midterm grade used for determining each student’s final grade is the average of all midterm quiz number grades, except that a separate average for each classroom group will be used if these individual classroom group averages differ by 10 points or more. This is done because of concerns that some classroom groups have more work experience or technical educational background than others. The students attending the recordings of the lecture form one classroom group. A group of remote learning students at a single organizational location, such as the XYZ Telecom Corp. typically forms another group. In most cases each group corresponds to a single SMU section number. Send the finished term paper, printed on paper , to the instructor or to Gary McCleskey at SMU, to arrive before Dec. 1, 2006 at 5PM. Late papers may temporarily receive an Incomplete grade until it is graded and corrections are put into the SMU ACCESS grade reporting system. In some cases, the instructor may request a computer-readable media like disk, after the paper is read and graded. It is not necessary to send such media unless specifically requested.
Many students in this course have come from educational backgrounds far away from the Electrical Engineering, Physics and Mathematics backgrounds that are traditionally associated with telecom technology. For example, at least two past students in the last 20 years had undergraduate degrees in forestry! This “non-technical” 70% of the students are typically employed in the telecom industry and need to understand the technological basis of their work, and for competent evaluation of assertions and claims made by equipment manufacturers (for service providers), and competent evaluation of alternative product development technologies (for manufacturers). Without the technological and economic “smarts” needed in the industry, significant resources have been devoted to projects that were not ultimately successful. Many observers of the telecom industry hold the opinion that the following are examples: Two that had competent technology but which did not succeed, due to prices higher than the end user was willing to pay, are the Iridium satellite radio network and in-flight radio-telephone service in scheduled airplane flights. A third example is ISDN, which was ultimately overtaken by lower cost improved high bit-rate modems (V.90 and V.92) These notes and this course supplement the equations and highly technical descriptions of various scientific and technological systems with a verbal and less quantitative description as well. We use a term paper (together with a one-hour midterm quiz) as the main grading factor for several reasons. A final examination (not used here) on technology details unfairly favors the students with a science or technology background, whereas each student can write a competent term paper at their own level of technological detail.
The Iridium satellite telecommunications system was originally designed by Motorola Corp. to provide complete world-wide portable digital voice and data communication, even in unpopulated areas where ordinary terrestrial cellular/wireless service was not expected to be available. The initial design used 77 LEO polar orbiting communication satellites, a number equal to the number of electrons in the Iridium atom – thus the product name. The number of satellites in the eventual working system was reduced to 66 as a cost saving change, still ensuring that at least one satellite was “visible” overhead at all locations. Although impressed by the advanced technology, many industry observers criticized the pricing when the plan was announced. Iridium and some other firms with planned LEO satellite systems placed advertisements proposing that impoverished farmers in remote areas of the globe would use LEO satellite service. Critics ridiculed these plans. Just as the critics expected, there were very few customers– typically a few oil-exploration crews who would 1) pay for high-cost service, and 2) wanted this service in remote un-populated areas . In most high and middle density populated areas, competing terrestrial wireless service was rapidly installed in the 1990s with low prices; typically $50 handsets and service charges of only pennies per minute. Ultimately, even extensive Iridium price cutting did not increase the user population sufficiently and the Iridium Corp. went into bankruptcy, still providing service to a few government agencies, but without enough capital to ultimately replace the LEO satellites expected to naturally burn up in the lower atmosphere in coming years.
Technically oriented people sometimes say, “We have the technology. Let’s build the product.” They sometimes do not undertake adequate market research to determine the anticipated cost of operation and planned income. Worse, they occasionally conduct badly planned market studies, or they ignore the predictions of well-planned marketing studies that they don’t want to hear. After these few introductory slides regarding non-technical aspects of product development, we will spend the rest of the semester on technical topics.
During the 1970s and 1980s, very significant amounts of telecom industry R&D resources were devoted to ISDN. Two major subscriber interfaces were developed and standardized: the Basic Rate Interface (BRI), providing one or two 64 kbit/s traffic (or Bearer “B”) channels, and the Primary Rate Interface (PRI), which in North America provides up to 23 B channels and one 64 kb/s data or “D” channel devoted to signals for call setup, disconnection, etc. There are a fairly significant number of PRI installations today, typically used with a PBX, . Some people point out that a PRI interface is physically identical with a T-1 digital multiplexer interface, and differs only with regard to the signaling messages in that D channel, and the related software, which conform to the Q.931 standard developed for ISDN. BRI installations are have the same customer target as ordinary analog telephone service:for residential and small business subscribers. Major ISDN hardware and software multi-billion dollar development costs in the 1980s were devoted to BRI, but BRI installations are few and far between today, even in European countries where the telephone service providers are allowed to cross-subsidize ISDN service from other revenue, and thus offer ISDN BRI at a lower price. In North America, BRI ISDN is significantly more costly in both installation and recurring costs than an analog telephone line fitted with a V.90 or V.92 modem, which can provide nearly equivalent “downlink” data rates to subscribers. In fact, the entire industry emphasis regarding high data bit rates for the public has shifted to ISDN, typically provided to subscribers via ADSL on telephone subscriber loops or via co-axial cable using cable TV installations. These two competitive technologies were not as fully developed in the 1980s and were not then viewed as serious competition to ISDN.
Supply-demand curves were first used in the late 19 th century by British economist Alfred Marshall, to explain the market mechanism that determines actual sales price. During a specified time interval, the behavior or motivation of the potential buyer (of candy bars, in Fig.A example) is theoretically described by a demand curve like D1-D2. This curve represents a particular buyer group who will buy a candy each day (30 per month) at a 50 cent price. If the price increases to $1, this average consumer under certain specific background circumstances, buys only 15 per month. If the price goes down to 25 cents, the average consumer buys 100 per month, and stores what he can’t eat in the freezer! This curve is typically constructed by interviewing potential consumers, or by observing the opinions and reaction of potential consumers in “focus groups.” For example, if a desirable substitute candy bar becomes available or Mr. Average goes on a diet, this particular curve no longer applies and fresh research and a new demand curve must be done. If advertising to increase the consumer’s desire to eat these candy bars is successful, a new demand curve, made by sliding the existing demand curve up and to the right, is required. The supply curve S1-S2 shows that the objective of the seller(s) (and manufacturer, etc. the “supply chain”) is to manufacture and sell more of most (not all) products when the sale price goes up. Also, when a certain product sells very well or is very profitable (example : bottled drinking water), competitive manufacturers typically enter the market with their equivalent product thus increasing the quantity available for a higher price. The point (typically only one on the diagram) where these two curves intersect is the equilibrium market price (point W in Fig. A). The product of the market unit price and the quantity sold (equal to the area of rectangle WXYZ) is the gross sales in dollars during the relevant time interval (a month in the example). If the seller’s costs (and thus the sales price) decrease due to new technology, reduced raw material costs or labor costs, or other reasons, a new supply curve made by sliding the existing supply curve to the left (and perhaps up as well), is used. Some products (like gasoline vehicle fuel) can’t be stored and the consumer typically must buy a fixed amount per month for essential driving (to and from work). The consumer must pay more if supply prices increase (supply curve shifts to the right). Illustrated in Fig. B. Some products are perceived as very low cost by consumers. For example, in Fig. C, consumers will thus not buy a telephone call priced over about 20 cents per minute from in-flight systems. The in-flight service provider has very high capital and operating costs, and thus prices the service very high when very few calls are made. In this case, the supply and demand curves intersect at a point where the unit sales quantity is zero, and gross income is thus zero.
Some systems are not successful because the quantity customers will buy, while not zero, is too small to give a profit at the price offered. Offering the product at a lower unit price increases the quantity sold, but in some instances the unit cost to manufacture the product is so high that a lower cost is prohibited. In contrast, some sellers price a new technology device (e.g. a new microprocessor chip) from the beginning at the same low price that is anticipated after it becomes a high quantity, high gross volume item. This encourages more consumers to buy it earlier, and thus reach high sales quantity from the start. This is called “forward pricing” or, with tongue in cheek, “self-fulfilling prophecy.” According to some pessimistic forecasters, direct subscriber-paid satellite radio broadcasting (Sirius and competitor XM Radio) is priced low, but is not profitable. Although both satellite radio systems are garnering significant publicity, and Sirius in particular has signed up so-called “shock-jock” radio personality Howard Stern, the number of subscribers to either system is significantly smaller than was forecast. Some of the consumer reluctance to buy subscriber-paid radio broadcasts is a mind-set based on “free” (sponsor-paid) service in the USA. Again, several other countries (such as Britain and Canada) have subscriber paid radio and TV broadcasting services (provided by a government-designated department), and its citizens would possibly be more receptive to satellite radio, but direct satellite radio service is not currently marketed outside of North America. Also, cable TV is a very similar consumer-paid service, and is successful. It is difficult to consistently and accurately extrapolate from one service to another. Ref. On Jack Goeken: www. goeken .com/ jack .htm
Aside from the customer's reaction to the product's price, there is also a possible issue with the billing or business practice regarding who pays and what “rights” are sold for continued use of a product. There are several relatively recent proposals whose lack of customer acceptance was ultimately identified (after products did not sell) as a perception by the potential customers that the proposed system was too far from the customer's concept of a fair system. Although this is not a technology issue, it often has a major technology impact because a wrong guess diverts scarce product development resources from making successful products.. In 1998-99, Circuit City retail electronics stores briefly marketed a system called DIVX, that required a special disk player and a small payment (credit card transaction via dial-up modem) each time the disk, already in the customer's physical possession, could be replayed. Similarly, Buena Vista (Disney studios) briefly marketed a “Flexplay” DVD that would darken within days after the air-tight package was opened, and thus “self destruct” (become unreadable by the optical DVD player). These two systems were proposed as alternatives to the historical method of renting video media, which then must be returned to the rental retail outlet. Although most preliminary market and focus group studies projected good customer acceptance, real customers rejected both systems, expressing their perception that once they pay for a disk, they should have the right to use it forever without further “per view” payments. The original industry concerns with disc returns and late fees seem to be addressed today via non-technical billing policy changes. References: http://www.cedmagic.com/history/rca-divx-ps8680z.html and http://www.wired.com/news/technology/0,1282,60983,00.html
Several standards organizations attempted to define a single world-wide compatible standard for 3G wireless, but they could not all agree. Ultimately, two incompatible standards based on different versions of CDMA technology were defined. Two major service providers in North America (Verizon and Sprint, who presently use CDMA, the most popular technology in North America), and a smaller contingent from several Asian countries, all favored CDMA-2000. GSM service providers in North America (Cingular, T-Mobile, etc.) and throughout the world (where it is the major technology) favored W-CDMA. Many observers attribute this to needlessly aggressive competitive. Some people in the industry hope that 3G technology with its “bells and whistles” (entertainment video images, etc.) will lead to higher consumer buying of cellular equipment and services, which has slowed in recent years, after continual growth in the last 10 years. The lowest revenue 2G voice service is the wireless market sector with the largest current (year 2006) growth rate. All 3G and 2.5G systems require new subscriber handsets for new services beyond traditional voice. Base equipment must be replaced or added to as well.
Some “3G sceptics” are very dubious about the cost and ultimate market size for 3G, regardless of specific technology. The sceptics accuse the 3G proponents of misdirecting development resources to an unmarketable product objective. Early 3G marketing trials in Japan and other Asian countries are not so promising. These “sceptics” developed so-called “2.5G” technology, obliged to use the “.5” or “1/2” terminology to fit it in between 2G and 3G.. 2.5G does not require such an expensive change to 2G base radio equipment, and the mobile handset is less complex than 3G, but it still provides a medium bit-rate packet data capability to existing 2G base systems. We will discuss GPRS and EDGE, two 2.5G technologies related to GSM, in a future lecture. There is also a small but growing possibility that the industry will ultimately skip over 3G CDMA in favor of a so-called 4G technology. In particular, OFDM, a technology with extremely high spectral efficiency, already used in ADSL, WiFi and other high bit rate digital systems, is promising. This is a complicated techno-economic question. If the decision was in your hands, which technology would you develop with limited funds?
In some cases the “fastest” (highest digital bit rate) wireless data system is not the best overall choice because the number of simultaneous users per cell or the cost of the system per conversation is not favorable. The ratio of the data bit rate (bits/second) of a system to the corresponding radio bandwidth (in Hz) is called the spectral efficiency of that system. A system having high spectral efficiency typically works satisfactorily (exhibiting a low BER) only when the ratio of signal power to the sum of noise and interference power is relatively large. This in turn requires that the system be operated with fewer simultaneous “conversations” in each cell. A better unit for comparing one cellular system with another is thus either: spectral geographic efficiency= number of conversations per cell · bit rate of a conversation/ radio bandwidth ·cell area . or, even more financially oriented: economic spectral-geographic efficiency = number of conversations per cell · bit rate of a conversation/ radio bandwidth · cell area · cost . Both equations refer to a “fully loaded” system at its maximum traffic level. The cost in the latter equation can be either equivalent capital present cost, or equivalent recurring cost (e.g., per month) of all the apparatus (and other operating expenses) involved in supporting one cell. Electrical “noise” is a fundamental undesired random variation or fluctuation in electric current or voltage, primarily due to the random thermal motion of electrons. When noise current waveforms are made audible via an earphone or loudspeaker the sound is like a hiss or like applause from a large audience or the sound of frying food on a stove. Electrical “interference” is caused by a man-made source that, in theory, can be turned off (for example, a neon sign). The major interference source in cellular radio systems is the radio signal transmitted by other radios that intentionally use the same radio frequency in a different geographic “cell.” Some authors occasionally use the term “noise” to cover both types of undesired signals.
Originator-pays-for-air-time in a wireless-destination call (with no cost to destination cell phone) was promoted in North America during the late 1990s as a major way to increase North American wireless subscriber population. Many people in the telephone industry believed that wireless subscriber population would increase substantially if subscribers did not have to pay “air time” for incoming calls. Despite glowing forecasts for such billing in the industry, it was almost totally rejected by originator callers in first marketing tests. It was never put in place permanently, despite the fact that much resources had been consumed to prepare and test the software and announcement system for this billing method. Curiously, many wireless networks outside North America traditionally use this billing method. Per minute or per-call pricing is also common for wired public telephone systems outside North America. To facilitate such billing when calls to a cell phone have a higher price, all wireless telephones in such a country are assigned one (or more) distinct wireless “area code(s)” (examples: 07- in the UK, 04- in Australia, 02- in New Zealand, 06- in Austria, France and the Netherlands, 05- in Israel, 015-, 016- and 017- in Germany, 9- in Ecuador, etc. )
Consumer electronic devices with multiple options or capabilities have been historically notorious for human interface complexity. The late comedian, Johnny Carson, confessed that he had a video cassette recorder (VCR) that was blinking “12:00” for the 9 years he owned it! He could not determine how to set the clock. The human interface of modern video recorders and players today are much improved, taking advantage of “on screen programming.” This is not true “computer programming,” but is the selection of the desired options from a sequence of menus displayed on the TV screen Many wireless telephone handsets today have many features and also have a sequence of menus that appear on a display screen to prompt the subscriber to set the options desired. Some wireless subscribers have complaints regarding the complexity of these menus, but several well-designed human interfaces are in use. Analogous to the user complexity issue in telecommunications, we remark that one of the things (aside from the obvious: high price!) that limits the growth of the “general aviation” market (for small aircraft typically owned by the pilot) is the complicated human interface between the aircraft and the pilot. The cockpit of even a small aircraft is stuffed with many different gauges and indicators (called a “clock shop”) that the pilot must continually monitor. Long, rigorous and costly training is needed to qualify for a pilot’s license and periodic medical examinations are required to keep the license. By comparison obtaining and keeping an automobile driver’s license is almost trivially simpler and quicker. Aircraft accidents are all investigated in detail and findings are used to improve aviation safety, while most automobile accidents ae not investigated at all. But when we consider that an automobile ride is 10 times more likely to result in fatality than an airplane journey of the same distance, some safety experts have suggested that driver licensing should be much more rigorous as well. In contrast, some aircraft marketers would like to have the requirements for piloting aircraft to be as fast and simple as an auto driver’s license, with the new generation of general aviation aircraft equipped with high-reliability computer control of every aspect of flight, to make air travel as safe and simple as can be. Aside from being an interesting point regarding human interfaces of technical systems, it also reflects how the public has become unconcerned about the high level of automobile traffic fatalities. Ref. For accidents/million passenger miles statistics: http://www.benbest.com/lifeext/causes.html
Focus groups and market trials are sometimes provide very accurate forecasts, but sometimes they are completely wrong. Some reasons for these discrepancies, like the “xxx effect” are understood by psychologists, but despite this knowledge many people in marketing focus groups say they will buy a product at a certain price, but in fact when given that real opportunity they won’t. Sometimes a good executive decision agrees with the formal product projections, and sometimes the opposite occurs as well. Focus groups and market trials sometimes provide very accurate forecasts, but sometimes they are completely wrong. Some reasons for these discrepancies, like the “ Hawthorne Effect, ” are well understood by industrial and marketing psychologists, but despite this knowledge many panel members in marketing focus groups say they will buy a product at a certain price, but in fact when given that real opportunity they won’t buy. The Hawthorne Effect was first described in the 1920s, when the Western Electric Company, then the manufacturing unit of the AT&T Bell System, brought in industrial psychologists to study the effect of ambient light level on the productivity of assembly line workers of its Hawthorne Works factory in suburban Chicago. The psychologists found that both increasing and decreasing the light level increased productivity. Neither change continued to increase productivity after weeks passed and the workers became accustomed to being studied and interviewed by the psychologists. The conclusion was that the workers were responding to the personal attention and not indicating their innate preferences. There is always a possible hazard with focus groups and other market projection methods involving contact, or even observation of the subject by the researcher that the Hawthorne Effect will distort the true forecast.
14 12 Sometimes a good executive decision agrees with the formal product projections, and sometimes the opposite occurs instead. It is very difficult to separate accurate from inaccurate business forecasts. When faced with forecasts that contradict common sense or each other, some executives call in additional consulting experts to advise them about future technology and products. Even then, different “experts” may contradict each other as well! Your instructor earns more income each year as a technological consultant than as a professor, but he still wants people in industry to understand how and why a technological conclusion is reached instead of merely choosing one consultant’s opinion over another. Therefore one objective of this course is to give each student enough scientific and technical background to allow them to research the subject and to form their own conclusions.
Many national telephone and postal systems are under the control of a government agency typically known as the department of Posts, Telegraph and Telephone (PTT). Although the objective of postal systems and telegraph-telephone systems is communication, the means and methods of the two kinds of telecommunication are so different that few resources are shared within such a PTT organization, and the telecom and postal systems have been separated in many countries with almost no negative effects. Electromagnetic waves guided by conductive metal wires are fundamentally the same as optical waves carried via fibers. They both involve rapidly alternating polarity of the electromagnetic field, but they differ in the number of oscillations per second. They differ so much that two quite distinct technologies have developed to use them, and they are now both used side by side in the telecom industry. It is interesting to note that electromagnetic waves have no “mass” or weight (Physicists say that photons, the “particle” model for electromagnetic waves, have “zero rest mass”). Some physicists have proposed (perhaps in jest) that telecommunication be carried via a beam of certain sub-atomic particles such as neutrinos. The proposed advantage of this is that neutrinos pass through everything. The great disadvantages are that extremely large and power-hungry apparatus is needed to produce neutrinos, and they are very difficult to detect (receive).
15 13 Historians report that Morse, a well-known artist and professor of art at New York University, learned from conversations with a physics professor about electric current.He conceived the idea of telecommunication by means of turning an electric current on and off in a time pattern of short (dot) and long (dash) intervals. The first version of his system printed the dots and dashes on a roll of paper tape, but technicians quickly realized that they could hear the sound patterns of each character and write them down manually.
An electro-mechanical repeater was a replacement for a hman repleater. A human telegrapher could write down the received message and then he or another teelgtapher cold repeat it by sending via the next link in the path from origin to destination. OF course, the human method delays the repeat of the message n most cases, and may introduce errors. From this historical beginning, we use the name “repeater” for any device that “regenerates” a stronger signal, retimes and reshapes the signal waveform; the classic functions of a repeater. Modern repeaters (particularly the Erbium-doped laser repeater used for certain optical fibers merely” amplifies (regenerates or makes stronger) the optical signal. Interesting historical note: Most counties in the state of Texas are approximately square, 30 miles on each side. This was done to make a 30 mile (50 km) telegraph link feasible from the center of one county to another in an adjacent county, without a repeater. Texas also required every county to build its county court house within 3 miles of the county center point. In the mid 19 th century many telegraph companies merged into the Western Union Corp., the first nationwide business and the first telecom monopoly.WU played an important role in the development of the telephone by not buying the patent when first offered. WU was bought and then sold by AT&T in the early 20 th century. WU discontinued telegram service in 2006 and now is primarily in the money transfer business.
15 13 Alexander Graham Bell and his father, Alexander Melville Bell, were both speech teachers specializing in teaching deaf people to speak understandably. During the 19 th century many diseases affecting the ear, such aas Scarlet Fever, were untreatable. A much larger percent of the population was deaf than today, although even today between 22 million and 24 million Americans are totally or partially deaf. Alexander Melville Bell invented a phonetic alphabet called “visible speech.” It used an “alphabet” of pictogram symbols that were mostly simplified cross-sectional drawings of the mouth and throat. When A. Graham Bell became wealth due to the income from the telephone, he subsidized having special moveable printing type custom made for these symbols and then for publishing the book Visible Speech , which was widely distributed (SMU’s library has it). It was not generally used by speech professionals, however. They have mostly used the International Phonetic Alphabet (IPA) invented at that same time by other speech experts. We will discuss it in a later lecture. Some people believe that Alexander Melville Bell is the living person that author George Bernard Shaw had in mind when he wrote of Prof. Henry Higgins in the book Pygmalion (later adapted by A.J.Lerner and F. Lowe into the musical play and film “My Fair Lady,” in which a speech teacher teaches a poor English girl with a lower-class accent to speak like a high-class lady, and thus makes her acceptable as a member of British high society. However, Shaw was well acquainted with several “dialecticians and grammarians,” and so some other historians say that Higgins is modeled on Daniel Jones, British author of a phonetic dictionary of the English language.
16 14 A.G.Bell studied the way sound, an oscillation fluctuation in local air pressure, was produced by the mouth. He realized that many sounds can be viewed as equivalent to the simultaneous presence of several sine waves at different frequencies. This was very similar to the behavior of a device he had invented to send simultaneous multiple telegraph signals over the same wire. Therefore he understood what was happening when he spoke near the set of vibratory metal reeds in his “transmitter” and receiver and the sound of his speech was heard in another room by his technician, Thomas Watson. Gardner G. Hubbard, father of one of Bell’s deaf students (and later Bell’s wife), was financing Bell’s development because Hubbard had a long term desire to compete with WU. Today we would say that Bell was trying to make a frequency division multiplexer for telegraph signal waveforms, and discovered that it could also transmit speech waveforms as well. A US patent comprises a “specification” that describes how the invention works in a “preferred implementation” and also several claims that describe more precisely what the inventor claims to have invented. After the patent application has been filed in the Patent Office, no change can be made to the specification, but typically the claims are modified/added/removed as part of the negotiation between the patent examiner and the inventor or his/her attorney.
17 15 Most historians agree that an offer to sell the Bell patent to WU was made, but none believe that the letter available on the web site is the true report of whatever group or committee WU assembled to evaluate the Bell Telephone patent. However, it has been a popular document since it shows just how wrong experts can be about the future of an invention. Note particularly that several of the complaints in this letter are absolutely correct, but later inventions by others have mostly corrected these problems or have reduced their severity.
18 16 WU was in a strong position with Edison’s microphone patent, but many historians conclude that WU still disliked Hubbard, and did not see long term big money coming from this telephone invention. Several historians have said that WU’s big mistake was not their refusal to buy the patent in 1876, but in failing to follow through by getting into the telephone business when they had the Edison microphone patent several years later.
19 17 Today it is illegal for to pay a firm or person not to compete with yourself, since the result will be an illegal monopoly. On the other hand, the owner of a valid patent has a time-limited (20 years) right to legally prevent others from making, using, selling, importing, etc. devices or services that use the patented invention. If others want to make, use, sell, etc., the patented product, they must make a mutually acceptable agreement, typically a payment of a relatively small “royalty” to the patent owner, on each unit sold.
20 18 The Kingsbury commitment was the first of several settlements between the Bell System (as At&t and its subsidiaries was long known) and the US Department of Justice, Anti-Trust Division.
21 19 This is a very abbreviated list. For many years in the 20 th century, Bell Laboratories, the central research and development division of AT&T filed more patent applications than any other company, although IBM took that prize some years.
Digital Telecommunications Technology - EETS8320 Fall 2006 <ul><li>Lecture 1 </li></ul><ul><li>Overview and Introduction </li></ul><ul><li>(Slides with notes.) </li></ul>
Introduction: EETS8320 <ul><li>Subject Area: Digital coding and multiplexing of telecommunications transmissions ( formerly in course EETS8302) </li></ul><ul><li>Digital telecommunications switching ( formerly in course EETS8304) </li></ul><ul><li>Descriptive and semi-technical treatment </li></ul><ul><ul><li>About 70% of our students do not have an engineering or science undergraduate degree, although many work in the telecom industry. </li></ul></ul><ul><ul><li>Each student can write a term paper at a technical level appropriate to their own background and knowledge </li></ul></ul>
Course Administrative Matters <ul><li>13 weeks of class each 3 hours (consecutive), with slides and notes </li></ul><ul><li>Each student takes a multiple-choice midterm quiz (1hour) and writes a term paper (approx 20 pages or 5000 words) on a pre-approved topic. The letter grade on the term paper is substantially your final grade. </li></ul><ul><li>If your midterm quiz numerical grade is above average* your final letter grade is increased by one “step” on SMU’s grade scale: Example, B+ A- </li></ul><ul><li>If your midterm is below average, no deduction is made. Your paper grade is then your course grade. </li></ul><ul><li>*Notes explain details. </li></ul>
Course Objectives <ul><li>One objective: to give students sufficient understanding of the technology to make intelligent decisions in the present and future </li></ul><ul><ul><li>This course is focused on science and technology, because understanding technology is important. </li></ul></ul><ul><ul><li>Adequate understanding of both technology and business are very important in the telecommunications industry. </li></ul></ul><ul><ul><li>Knowledge of business-economics alone is not sufficient! </li></ul></ul><ul><ul><li>Knowledge of technology alone, with ignorance of economics is also not sufficient! </li></ul></ul><ul><ul><ul><li>The Iridium system, ISDN and In-flight telephones are un-successful telecom products often cited as examples of economic or technological bad judgement. </li></ul></ul></ul><ul><ul><li>Human interface (ease of use) is also a factor in some cases. </li></ul></ul>
Problem Products: Iridium <ul><li>Iridium, a world-wide direct satellite telecom system of 1990s </li></ul><ul><li>Technologically impressive, but… </li></ul><ul><ul><li>Priced higher than most potential customers would pay: </li></ul></ul><ul><ul><ul><li>Handset $3000 (price later significantly reduced) </li></ul></ul></ul><ul><ul><ul><li>Service $3/min or more (price later slightly reduced) </li></ul></ul></ul><ul><ul><ul><li>Designers and implementers were aware of possible low sales risk due to high prices. </li></ul></ul></ul><ul><ul><li>Unexpected low cost terrestrial competition in populated areas harmed Iridum . </li></ul></ul>
Iridium: More Recent History <ul><li>High-budget Customer Base was never large enough: </li></ul><ul><ul><li>For example, oil exploration crews in Siberia? Very few of these! </li></ul></ul><ul><ul><li>Native farmers in Kazakhstan? They could not afford Iridium. </li></ul></ul><ul><ul><li>Callers from an ocean liner? Sounds promising: </li></ul></ul><ul><ul><ul><li>Existing Inmarsat satellite telephone calls are $10/minute! </li></ul></ul></ul><ul><ul><ul><li>But e-mail to/from most ocean liners is free!! </li></ul></ul></ul><ul><li>If you build it, will they come?… Apparently, No! </li></ul><ul><ul><li>Customer enrolment was only a tiny fraction of Iridium management’s estimates. Several top Iridium executives resigned. </li></ul></ul><ul><ul><li>Iridium Corp. filed for Chapter 11 bankruptcy protection in August, 1999. This allows continued operation while a plan is made to hopefully reorganize and eventually pay creditors (bondholders, etc.). Shareholders are not protected in Chapter 11. Reorganized as Iridium Satellite LLC in Dec. 2000 </li></ul></ul><ul><ul><li>US Government subscribers are almost the only present users, while Iridium operates in reorganization. </li></ul></ul><ul><ul><li>Development of several competitive LEO satellite systems (e.g. Globalstar) stopped. Licenses were cancelled or returned to the FCC. </li></ul></ul>
Problem Products: ISDN <ul><li>Fully digital end-to-end telecommunication via 64 kbit/s channels derived from pre-existing digital telecom channels </li></ul><ul><li>Was viewed as the unquestioned future direction of PSTN voice-data service in the 1980s. </li></ul><ul><ul><li>Bone of contention among major telephone switch manufacturers. </li></ul></ul><ul><li>Ultimately in limited use but not in consumer demand due to high cost </li></ul><ul><ul><li>Unanticipated availability of low-cost 53 kbit/s V.90 modems in 1990s diverted many potential customers </li></ul></ul>
A Page from Economic Theory <ul><li>Supply-demand curves graphically describe how buyers and sellers find an equilibrium price per unit and quantity sold. See notes. </li></ul>Quantity bought or sold Unit price ($/gal) Quantity bought or sold 0.50 1.00 50 100 Quantity minutes bought or sold 0.50 1.00 50 100 Unit price ($) Unit price/min ($) A B C . W .25 15 30 X Z Y D1 D2 S2 S1 . In-flight Phone Service. Motor Fuel. 30 . Quantity bought or sold 5.00 10.00 50 100
Problem Products: In-Flight Telephone Service <ul><li>Very costly to install due to severe aircraft radio interference standards. </li></ul><ul><ul><li>Originally allowed only aircraft-originated calls. Aircraft destination calls are supported in some systems in a somewhat inconvenient way.. </li></ul></ul><ul><li>None were ever profitable. Many competitive systems addressed improved convenience, video games, etc., but not low price. </li></ul><ul><ul><li>Test market studies show that sales improve dramatically below 20 cents/min, but existing systems can't meet that price. </li></ul></ul><ul><li>Verizon Airfone plans, after 21 years, to discontinue service by the end of 2006. In-flight phone service will end on over 1,000 aircraft operated by United, Delta, Continental, US Airways, Air Canada and Cathay Pacific. </li></ul><ul><ul><li>Proposed systems using customers' existing cell phones inside the aircraft are still in preliminary development. </li></ul></ul><ul><ul><li>Use of VoIP via Internet links (priced at ~$10/h)) provided in some flights is another alternative </li></ul></ul><ul><li>Historical note: First (analog) Airphone system developed by Jack Goeken, also famed as founder of MCI . </li></ul>
Customer Perception of “Fairness” is Important <ul><li>Some system proposals did not succeed due to negative customer perception of “fairness” </li></ul><ul><ul><li>Two types of limited play video disks were test marketed circa 1998 as “no return” methods for video rentals. Both rejected by customers. </li></ul></ul><ul><ul><li>System software for wireless air time charges paid by land-line originator were developed, due to industry pressure circa 2000, but 100% of participants in US marketing tests would not choose this billing method. </li></ul></ul>
Concerns about 3G Wireless <ul><li>Some telecom industry observers fear that 3G, and other advanced wireless data technologies, will suffer fates like those of Iridium and ISDN. </li></ul><ul><ul><li>First Generation (1G) wireless was analog cellular technology used from1981 to mid 1990s. Very few still in use. </li></ul></ul><ul><ul><li>Second Generation (2G) wireless utilizes digital speech coding, used from early 1990s to the present. Technologies include GSM, CDMA, North American TDMA, iDEN (NexTel) and others. Range of available bit rate per user is about 6 kbit/s to 20 kbit/s </li></ul></ul><ul><ul><li>Two and a Half (2.5G or 2-1/2G) designs are packet data systems, achieving available bit rate per user up to about 60 kbit/s to 140 kbit/s. Used for Internet access and packet voice (VoIP). </li></ul></ul><ul><ul><li>Third Generation (3G) utilizes various types of CDMA (UMTS, CDMA2000). Provides bit rates up to 2 Mbit/s. Major applications are viewing high quality visual entertainment (HDTV images) or possibly transferring massive files via Internet. </li></ul></ul><ul><ul><li>Fourth Generation (4G) utilizes OFDM to achieve 16 Mbit/s or higher bit rates. Applications similar to 3G but even “faster.” </li></ul></ul>
2.5 G <ul><li>These “3G sceptics” believe that the major growth in cellular industry will come from lower cost voice service and less glamorous data services like e-mail. </li></ul><ul><ul><li>They therefore designed a packet data technology upgrade based on GSM, called GPRS (and a higher data rate version named EDGE), called 2.5G, or “2 and a half G.” </li></ul></ul><ul><ul><li>The cost of installing GPRS or EDGE in an existing GSM base system is relatively small. (In contrast, all 3G systems require costly new or additional base radio replacements.) </li></ul></ul><ul><ul><ul><li>GPRS upgrades are already in place in some countries in Europe, N.America, Asia, Australia, etc. </li></ul></ul></ul><ul><ul><ul><li>Use of GPRS (later EDGE) in USA (by AT&T and Cingular) to replace IS-136 TDMA is under way today. Voice Stream (T-Mobile) has a GSM starting point technology and thus a less costly upgrade to GPRS and EDGE. Merger rumors between GSM-technology firms are continually rife. </li></ul></ul></ul><ul><ul><li>GPRS provides up to 171 kbit/s per subscriber, EDGE up to 384 kbit/s. </li></ul></ul>
Other Aspects <ul><li>Previous slides did not analyze or compare: </li></ul><ul><ul><li>Radio bandwidth required for higher bit rates. </li></ul></ul><ul><ul><li>Sensitivity of each technology to “noise” and interference limits the number of simultaneous “conversations” in each cell and thus the system capacity </li></ul></ul><ul><ul><li>Cost and complexity of each technology </li></ul></ul><ul><ul><li>Power consumption in active and standby modes, affecting battery “life.” </li></ul></ul><ul><ul><li>An accumulation of negative aspects like these can severely degrade the theoretical performance of real systems </li></ul></ul>
Customer Preference Issues <ul><li>In some cases, potential customers won't buy because they perceive cost or terms of sale inherently unattractive. Examples: </li></ul><ul><li>Today North American wireless subscribers view air time costs over about US$ 0.20/min as excessive </li></ul><ul><li>North American telephone users reject “caller pays” for wireless destination calls (although this is accepted in many other countries). </li></ul><ul><li>In a non-telecom case, customers reject a “self destructing” or “pay per view” video disc. </li></ul>
Sometimes Non-Technical Problems Dominate <ul><li>“ Morse code” telegraph was not practical for most end users because of the special skill required to send with a “key” and receive by listening to “di-dahs” </li></ul><ul><ul><li>Electro-mechanical Teletypewriter machines only require the ability to read and type (keyboard) but they were costly, bulky and noisy. </li></ul></ul><ul><ul><li>Telephone station sets were always relatively small, quiet when idle, and require only the ability to speak and hear understand the language of the other person. </li></ul></ul><ul><ul><ul><li>Special teletypewriters are available for deaf or hard of hearing telephone users. </li></ul></ul></ul>
How Does Customer Perceive Acceptable Price vs. Performance? <ul><li>In some cases, the technical performance is adequate, but end users perceive the price as excessive and won’t buy the product. </li></ul><ul><li>This non-technical aspect of product development is supposed to be addressed by customer surveys, focus groups, etc., but sometimes they predict incorrectly. </li></ul><ul><li>Telecom items perceived as overpriced: </li></ul><ul><ul><li>Iridium originally charged $3000 for a handset and $3 per minute air time </li></ul></ul><ul><ul><li>Scheduled airline in-flight telephones. Those systems still charged at least $2 or more per minute,due to high operating costs. </li></ul></ul><ul><li>Most end users (apparently) won’t pay over about $0.10 to $0.20/minute for “air time” </li></ul>
Best to Understand Technology Yourself <ul><li>Make well-founded decisions yourself </li></ul><ul><ul><li>Less dependence on the opinions of others </li></ul></ul><ul><ul><li>Your instructor earns most of his income from being a “technology expert” consultant, but would still rather have his clients understand the technology themselves! </li></ul></ul><ul><li>Separate the wheat from the chaff when exaggerated product claims are made </li></ul><ul><li>Make realistic and profitable product and service plans </li></ul><ul><ul><li>Do customers exist for this product or service? </li></ul></ul><ul><ul><li>Are they willing to pay a compensatory price for the product at projected costs? </li></ul></ul><ul><ul><li>Why is the product competitively advantageous? What are the competitive products or services? </li></ul></ul>
Now: Telecom Technology <ul><li>Having said enough for now about the reasons and motivations for telecom products, we turn to the technology of telecommunication. </li></ul><ul><li>There are two ways to convey information: </li></ul><ul><ul><li>Send a physical object. Historically, the customary object is a letter (e.g. on papyrus, parchment and later on paper). </li></ul></ul><ul><ul><li>Send some “energy” in the form of an electromagnetic wave. In ancient times, light was involved in viewing signals or semaphore signals at a distance. Privacy and data rate improvements had to wait for the discovery and a minimal understanding of electricity. </li></ul></ul>
Historical Overview: Telegraph <ul><li>Invented by Samuel F.B.Morse (an artist, not a scientist) greatly assisted by Alfred Vail* . </li></ul><ul><ul><li>Inter-city telegraph demonstrated by Morse in 1837. </li></ul></ul><ul><ul><li>Several less practical European telegraph systems preceded Morse </li></ul></ul><ul><li>For example, Morse (and others) thought that electrical signals travelled “instantaneously” from telegraph key to the sounder (receiver), since the complete theory of electromagnetic waves was not formulated until 1860-90 by J.C. Maxwell, O. Heaviside, et al. </li></ul><ul><li>* Coincidentally, a relative of Theodore Vail, president of AT&T about 60 years later </li></ul>
Telegraph Main Features <ul><li>Current flow around a circuit including a battery, telegraph key (on-off switch), a single wire (typically iron, later copper) with the earth as a return path. </li></ul><ul><li>Worked adequately up to about 30 miles, depending on earth conductivity. </li></ul><ul><li>About 1849 the repeater allowed longer links by chaining 30 mi sections via an electro-mechanical “relay” (switching contacts operated by an electromagnet). </li></ul>
Telecom Overview: Telephone <ul><li>The telephone was invented in 1876 by Alexander G. Bell (a speech teacher, not a scientist). Born in Scotland, Bell immigrated to Canada and then the USA. </li></ul><ul><li>The telephone had the significant advantage that no special skill (such as learning Morse code) was required to use it! </li></ul><ul><ul><li>Ease or convenience of use is often a deciding factor in the success of one technology over another. </li></ul></ul><ul><li>Bell’s microphone (called “transmitter”) produced electric current proportional to instantaneous air pressure. Earphone (“receiver”) reversed the process, converting the electrical waveform back into acoustic (sound) form. </li></ul>
Some Business History <ul><li>Bell was financed by his wealthy industrialist father-in-law, Gardiner G. Hubbard, a man with a history of business and legal contention with the (then) large Western Union Telegraph Company </li></ul><ul><li>Bell’s original objective was to send several independent telegraph signals over the same circuit </li></ul><ul><ul><li>Today we would describe his plan as frequency division multiplexing (FDM) of amplitude modulated Morse code. </li></ul></ul><ul><li>He discovered by accident that his equipment could transmit speech </li></ul><ul><ul><li>He added a new claim to his already filed patent covering this </li></ul></ul><ul><ul><li>When the telephone became commercially important, major patent litigation followed, ultimately decided by the US Supreme Court </li></ul></ul>
Business Conflicts <ul><li>Bell and Hubbard offered the patent to Western Union (WU) at first for $100,000 </li></ul><ul><ul><li>This was an immense sum* in 1876, when a large house cost less than $1000. </li></ul></ul><ul><ul><li>WU turned them down, due to dislike of Hubbard </li></ul></ul><ul><ul><ul><li>A famous negative evaluation letter (probably not authentic) is available on this course web site. </li></ul></ul></ul><ul><ul><ul><li>The letter also is a prime example of </li></ul></ul></ul><ul><ul><ul><ul><li>“ Not Invented Here” (NIH) attitude, ignoring good outside ideas </li></ul></ul></ul></ul><ul><ul><ul><ul><li>Lack of proper appreciation of the advantages of the invention </li></ul></ul></ul></ul><ul><ul><ul><ul><li>Inability to accurately foresee that improvements are possible to overcome the initial disadvantages of the invention </li></ul></ul></ul></ul><ul><ul><li>*The Internet web site http://eh.net/ehresources/howmuch/dollarq.php that contains a history of US dollar inflation, indicates that $100,000 in 1876 had the purchasing power of $1,705,922.48 in the year 2005. </li></ul></ul>
Early Competitive Moves <ul><li>WU, after recognizing the fast growth of the telephone, quickly decided to get back into competition </li></ul><ul><li>They hired the best available inventor, Thomas A. Edison, to invent a significantly improved microphone circa 1878 </li></ul><ul><ul><li>Edison studied the telephone, found its most important weakness, and came up with a solution based on Bell’s “liquid transmitter.” Bell’s liquid transmitter was a variable resistance microphone used in his first working voice transmission, but it was impractical because it used an acid-water solution as the variable resistance material. Edison substituted a sealed capsule of powdered carbon as the pressure sensitive variable resistance element. This “carbon microphone” invention was also later improved by German-American Emil Berliner as well. </li></ul></ul>
Business Strategies <ul><li>The improvement in audio loudness (permitting longer telephone wires and thus more wire coverage area per central office) gave the carbon microphone a strong economic competitive advantage. </li></ul><ul><li>But WU could not operate a telephone system without infringing the basic Bell patent, either. </li></ul><ul><ul><li>Negotiations were stalled, until the Bell company suggested something which would be illegal under present anti-trust law… </li></ul></ul><ul><li>WU agreed in 1879 to stay out of the telephone business for 20 years in return for 1% of the income from the telephone. </li></ul><ul><ul><li>The telephone industry grew so fast that Bell was soon able to buy most of WU shares. WU became a subsidiary of Bell from circa 1900 until divested in a famous 1914 anti-trust case. </li></ul></ul>
Early 20th Century <ul><li>American Telephone & Telegraph (the renamed Bell Telephone company), was headed for many years by Theodore Vail, coincidentally a nephew of Alfred Vail, Morse’s collaborator. </li></ul><ul><li>Vail vigorously bought out other telephone operating companies in most major cities, leaving only rural areas to the independents (formed after the Bell patents expired). This acquisition stopped in 1914*. </li></ul><ul><li>AT&T purchased Western Electric Co. (electric equipment manufacturer originally so named to save the cost of repainting the entire sign in a former Western Union repair shop), “vertically integrating” manufacturing and telephone operations </li></ul><ul><li>AT&T established its Long Lines division, providing long distance connection between all North American and foreign cities. </li></ul><ul><li>*In 1914 AT&T’s negotiator made the “Kingsbury commitment” to not buy out any more independent telephone companies, thus settling a federal antitrust lawsuit. </li></ul>
Some Technological Transmission Advances <ul><li>Single wire with earth return was replaced in 1890s by a subscriber “loop” of current carrying copper wire. </li></ul><ul><ul><li>An innovation by J.J.Carty, who became head of AT&T R&D and ultimately established Bell Telephone Laboratories. </li></ul></ul><ul><li>Some AT&T accomplishments during the first half of the 20th century: </li></ul><ul><li>DeForest’s “Audion” triode vacuum tube amplifier was improved and adapted for analog voice frequency amplification, leading to coast to coast long distance telephone connections. </li></ul><ul><li>Gilbert S. Vernam invented the Vernam Cipher cryptography method for teletypewriters during WW 1 </li></ul><ul><li>The quality and noise of analog telephone connections were improved in 1920s by H.S.Black’s invention of “negative feedback” at Bell Laboratories. </li></ul><ul><li>Frequency Division Multiplexing (FDM) using single side band (SSB) modulation was developed by John R.Carson at Bell Laboratories. Basis of Analog telephone multiplexing. </li></ul><ul><li>Microwave co-ax cable was developed by Lloyd Espenscheid of Bell Labs. Used today for T-3 and other signals. </li></ul>
More Business Developments <ul><li>The Anti-trust Division of the US Justice Department investigated AT&T in 1914, 1937, 1948, 1965, 1972. Each investigation led to consensual settlements which further restricted the scope of AT&Ts business. </li></ul><ul><ul><li>1914: “Kingsbury Commitment” stopped acquisition of independent telephone companies and divested WU from AT&T </li></ul></ul><ul><ul><li>1937: AT&T divested non-telephone businesses (appliances, motion picture sound systems, etc.) and offshore manufacturing. </li></ul></ul><ul><ul><ul><li>ITT (originally International Telephone and Telegraph Corp.) was founded by brothers Sosthenes and Hernand Behn. They were sugar brokers in Porto Rico who first bought the Puerto Rico telephone company. Then they founded Cia. Telefonica Espana in 1923. In 1937, using J.P. Morgan funds, they bought all off-shore Western Electric factories. They later founded other telephone companies in Latin America. ITT sold its telephone manufacturing businesses in 1990s to Alcatel, and now owns hotels, insurance companies and some non-telephone manufacturing firms. </li></ul></ul></ul><ul><ul><li>1948: AT&T agreed to license all patents to competitors </li></ul></ul><ul><ul><li>1969: AT&T agreed to allow connection of customer-owned equipment (result of FCC and court CarterPhone decision) rather than renting. AT&T had previously always rented equipment to the subscriber, a method learned from the United Shoe Machinery company in the early Boston years. </li></ul></ul><ul><ul><li>1984: AT&T divested local telcos (RBOCs) but retained long distance and manufacturing. (Manufacturing later separated under the Lucent name.) </li></ul></ul>
Other Business Events <ul><li>AT&T, until 1984 divestiture, received 1% of gross income of all RBOCs </li></ul><ul><ul><li>Also was part owner of Bell Canada and Northern Electric, its manufacturing subsidiary, until 1970s. This became Nortel Networks, no longer owned by AT&T. </li></ul></ul><ul><ul><li>Extensive cross-licensing of patents with other major telephone equipment manufacturers in other countries as well. </li></ul></ul><ul><ul><ul><li>Example: Crossbar telephone switch was developed under cross-licensing agreement with L.M.Ericsson of Sweden </li></ul></ul></ul><ul><li>AT&T acquired NCR (formerly National Cash Register) in 1989, then spun it off as part of the 1996 separation into three businesses. Lucent (with Bell Laboratories) is only a manufacturer and recently merged with Alcatel. AT&T is today an operating company in long distance. Its cellular/PCS activity is a separate corporation, now merged with Cingular Wireless. </li></ul><ul><ul><li>Both AT&T and Lucent have started several spin-offs also </li></ul></ul>
Some Major Telecom Vendors with Dallas-Ft.Worth Presence <ul><li>Alcatel (France) acquired most ITT manufacturing operations and Rockwell (Collins) telecom products, and Digital Switch Corp. (DSC), and integrated them with its existing products in the 1980-90s. </li></ul><ul><li>Ericsson (Sweden), another long term telecom manufacturer worldwide, has operations here. </li></ul><ul><li>Fujitsu, NEC (Nippon Electric Co.) are two separate independent Japanese telecom manufacturers with Dallas area operations </li></ul><ul><li>Motorola, primarily in Fort Worth (cellular and paging equipment) </li></ul><ul><li>Nokia (Finland), strong in cellular/PCS handsets but also makes cellular infrastructure and landline telecom switchgear </li></ul><ul><li>Nortel Networks (formerly Northern Telecom) is a descendant of Northern Electric of Canada. </li></ul><ul><li>Siemens (Germany) a long term telecom and general electrical equipment maker, now reducing its presence in telecom. </li></ul><ul><li>This area is sometimes called “Telecom Corridor ™ ” or “Switch Alley” </li></ul>
Some Telephone Operating Companies <ul><li>Originally 7 (now 4) Regional RBOCs, with consolidation of SWBell-PacTel-SNET and NYNEX-GTE-Bell Atlantic (now Verizon), etc. </li></ul><ul><li>GTE, arising from mid-century consolidation of many independent local telcos, merged with Bell Atlantic and Primeco wireless to form Verizon (rhymes with horizon) in 2000 </li></ul><ul><li>Scattered remaining independents in some smaller cities (e.g. Rochester NY, etc.) </li></ul><ul><li>Numerous Inter-Exchange Carriers (IXCs) the largest 3 being AT&T, MCI and Sprint. </li></ul><ul><li>The government operates Post, Telephone and Telegraph (PTT) administrations in many other countries; but “privatization” is spreading rapidly </li></ul>
Digital Telecom Revolution <ul><li>The T-1* digital multiplexing system, introduced by Bell Labs in 1961, ultimately led to an almost complete conversion of the North American public switched telephone network (PSTN) to digital transmission and (later) digital switching </li></ul><ul><li>T-1 was a rare and uniquely successful product because it is: </li></ul><ul><ul><li>Immediately equal or lower in cost than the prior analog FDM multiplexer. Cost improved more later with product evolution as well. </li></ul></ul><ul><ul><li>Carefully designed to be backwards compatible with all switching and prior art transmission equipment at connection interfaces </li></ul></ul><ul><ul><li>Better signal quality than FDM multiplex </li></ul></ul><ul><ul><li>More capacity (24 voice channels on the same wires that previously carried only 12 channels) </li></ul></ul><ul><li>*Also written T1. Since T-1 is a trade name, DS-1 is an approximately equivalent term used in standards documents, etc. </li></ul>
Is “Digital” Always Better? <ul><li>The “error” introduced by conversion from analog to digital representation can be controlled and limited in advance by the designer of the A/D converter. Called “quantizing” error. </li></ul><ul><li>Digital representation of information does not suffer from cumulative “noise” errors. Transmission over a longer distance only causes time delay, not distortion. </li></ul><ul><ul><li>This is the result of a system design in which the two binary digital voltage levels (typically 0 and 5 volts) differ by much more than the typical “noise” voltage level (typically 0.001 volts). </li></ul></ul><ul><li>But… digital representation typically uses more (radio) bandwidth than analog representations. </li></ul><ul><ul><li>This problem can be reduced by use of data compression coding in some cases. </li></ul></ul><ul><ul><li>When the channel is extremely “noisy,” (e.g., a cellular radio link) error protection coding must be used, and this requires part of the total channel bit rate to be devoted to bits for this purpose. Cellular radio systems typically devote half the physical bit rate capacity to error protection. </li></ul></ul>
T-1 Benefited From Prior Technology <ul><li>PSTN voice signals were historically already low-pass audio-frequency filtered to attenuate audio power above approx. 3.5 kHz audio frequency </li></ul><ul><ul><li>Necessary for FDM multiplexing and well-verified to support intelligible conversation </li></ul></ul><ul><ul><li>Permits accurate digital waveform coding at 8000 samples/second </li></ul></ul><ul><li>T-1 design was an early application for transistors </li></ul><ul><ul><li>Repeaters are installed at 6000 ft. intervals in outdoor or difficult-access locations and must operate reliably and consume little power </li></ul></ul><ul><ul><li>Vacuum tube devices would not be practical </li></ul></ul><ul><li>T-1 uses PCM (pulse code modulation) waveform coding with logarithmic companding </li></ul><ul><ul><li>8-bit binary coding of each waveform sample, with non-uniform voltage steps, produces uniform signal to noise ratio over a wide range of audio loudness </li></ul></ul><ul><li>8 bit/sample • 8000 sample/sec = 64,000 bit/second = 64 kb/s </li></ul>
Incorporation of Call-Processing Signals <ul><li>Two methods for signaling are in general North American use </li></ul><ul><ul><li>1. “Robbed bit” signaling uses the least significant bit of the PCM in every 6th frame to convey supervision (channel busy/idle) status. Five of every six consecutive waveform samples are not affected. </li></ul></ul><ul><ul><li>Systems for 12 and 24 multi-frame synchronizing patterns are used to ensure that the signaling equipment uses the proper bit </li></ul></ul><ul><ul><ul><li>Robbed bit signaling leaves 56 kb/s (7 bits of every sample) for the subscriber, even if not multi-frame synchronized </li></ul></ul></ul><ul><ul><li>2. Common channel signaling uses a reserved digital channel (either 64 or 1536 kb/s in North America) to convey messages in packet data form between switching systems regarding the call processing on numerous other channels </li></ul></ul><ul><ul><li>Common channel signaling system Number 7 is today’s world-wide standard, with some national variants There are many abbreviations: (SS7, CCS7, etc.) In some cases, different abbreviations imply different national variants of Common Channel Number 7. </li></ul></ul>
Further Digital Multiplexing <ul><li>Higher level digital multiplexing systems were developed with better economy for high traffic corridors: </li></ul><ul><ul><li>T-1 (DS-1): so called North American (and Japan) Primary Rate digital multiplexing. 24 channels at 1.544 Mb/s </li></ul></ul><ul><ul><li>T-1C: a double capacity system (48 channels) now rarely used. Not mentioned in international standards. 3.152 Mb/s </li></ul></ul><ul><ul><li>T-2 (DS-2): a quadruple capacity system (96 channels). Called M12 or Secondary Rate. Combines 4 DS-1 tributaries. Seldom installed today. 6.312 Mb/s </li></ul></ul><ul><ul><li>T-3 (DS-3): Combines 7 DS-2 tributaries. M13 multiplexers produce this Tertiary level rate from 28 T-1 tributaries. 44.736 Mb/s. Uses co-axial cable or microwaves. </li></ul></ul><ul><ul><li>Different and mostly incompatible “T-4” higher level digital multiplexers using co-axial cable or microwaves were developed in different countries (US, Canada, Japan) but were relatively little used since only a few routes have enough traffic to make this economically feasible. </li></ul></ul><ul><ul><li>European digital multiplexers of similar characteristics are widely used in other countries. </li></ul></ul>
Higher Level Multiplexer Trends <ul><li>DS-1, DS-2, DS-3 multiplexers are designed to accommodate small time-varying inaccuracy in the bit rate of the incoming tributaries (plesiochronous multiplexing) </li></ul><ul><li>An undesirably large portion of the total bit rate (bit “overhead”) is needed to handle this, and the necessary process for de-multiplexing a single DS-1 or single 64 kb/s voice channel (DS-0) is complicated and costly </li></ul><ul><li>These difficulties, and the increased use of fiber optic transmission and more accurate digital bit stream synchronization, has led to development of new and fundamentally improved multiplexing designs </li></ul>
EM Wave Transmission Media <ul><li>Radio transmission. Non-guided via open space </li></ul><ul><ul><li>Inferior channel characteristics due to fading, interference, etc. </li></ul></ul><ul><ul><li>But portability makes cellular service valuable, and absence of intermediate equipment between microwave towers gives lower cost. </li></ul></ul><ul><li>Guided electromagnetic waves: </li></ul><ul><ul><li>Via twisted pair wires, co-axial cable. Typically using “repeaters” to compensate for signal loss </li></ul></ul><ul><ul><li>Via optical fiber, in the infra-red optical frequency range. Electro-optical or all-optical signal amplifiers are used to compensate for losses. </li></ul></ul>
SONET and SDH <ul><li>A multiplexing format normally used on optical fiber, but lowest bit rate members of the family can be transmitted via co-axial cable or microwave radio. </li></ul><ul><ul><li>SONET (Synchronous Optical Network) in North America </li></ul></ul><ul><ul><li>SDH (Synchronous Digital Hierarchy) elsewhere </li></ul></ul><ul><li>In a refreshing departure from previous international incompatibility, these standards are virtually identical. SONET includes a lowest bit rate version at 51.84 Mb/s which is not used in SDH, but higher rates such as 155.52 Mb/s etc. are common to both standards and are compatible when similarly configured. </li></ul>
Digital Transmission and Switching <ul><li>The rapid growth of digital multiplexing transmission systems (almost 100% of the North American network today) led to a parallel development of digital local and long distance switches. These switches are more compact, use less power, and are more reliable than their electro-mechanical predecessors, and mostly contain automatic self-test equipment to permit efficient use of fewer repair personnel </li></ul><ul><li>Digital Switching is now included in the present course EETS8320. </li></ul>
Digital Switch Basics <ul><li>End office digital switches typically support traditional analog telephone sets, and in some cases ISDN or proprietary digital telephone sets. The analog voice waveform voltage is periodically measured (“sampled”) and each voltage is converted via analog/digital converters and digitally coded into a bit stream. </li></ul><ul><li>From trunk connections, separate channels of digital information are separated from the bit streams. </li></ul><ul><li>Digital channel data is stored temporarily (typically for 125 microseconds) in a local memory in the switch. </li></ul><ul><li>Desired outgoing channel bit streams are multiplexed together to connect to other switches. </li></ul><ul><li>Microprocessor internal to the switch controls routing of connections. </li></ul>
Switch Types <ul><li>End-office switch: both trunks and telephone sets </li></ul><ul><ul><li>Formerly called Class 5 </li></ul></ul><ul><li>Trunk-trunk switch (no telephone sets) </li></ul><ul><ul><li>Used to complete long distance connections between end switches </li></ul></ul><ul><ul><li>Used (with radio base stations) for cellular radio systems </li></ul></ul><ul><ul><li>Formerly designated as Class1 to Class 4 based on details of application in the network. </li></ul></ul><ul><li>Private Branch Exchange (PBX) switches, used primarily for business users to establish both external and internal calls </li></ul><ul><li>“ Intercom” switch. Connects telephone sets or “hands-free” stations for internal calls only. Rare today. </li></ul>
Switch Features <ul><li>All digital switches are microprocessor controlled and have many features. Some examples: </li></ul><ul><ul><li>Call waiting (signal during a conversation that another caller is attempting to reach you, and ability to answer that caller) </li></ul></ul><ul><ul><li>Incoming call forwarding to another number when desired </li></ul></ul><ul><ul><li>3-way conference calling via a conference bridge </li></ul></ul><ul><li>PBXs in particular have a large repertoire of sophisticated features. </li></ul>
Digital Speech Coding <ul><li>A technical “race” has continued for the last quarter century between speech coding technology and transmission technology </li></ul><ul><ul><li>Lower bit rate speech coders are typically more complex devices, but they allow carrying more conversations in a transmission medium with a fixed total bit rate </li></ul></ul><ul><ul><li>Innovations such as fiber optic transmission and integrated circuits have reduced the cost of high transmission bit rates </li></ul></ul><ul><li>The public telephone industry almost changed over to 32 kb/s ADPCM* speech coding in the early 1980s, but the lower cost of fiber stopped this plan </li></ul><ul><li>Radio systems such as cellular and PCS appear to be the main present use for lower bit rate speech coders. </li></ul><ul><li>* Adaptive Differential PCM </li></ul>
Other Speech Coding <ul><li>Digital speech coding methods generally fall into one of two categories: </li></ul><ul><li>1. Waveform coding. Examples include: </li></ul><ul><ul><li>PCM (Mu-law and A-law pulse code modulation: used in DS-1 and E-1) </li></ul></ul><ul><ul><li>ADPCM (adaptive differential PCM - typically 32 kb/s) </li></ul></ul><ul><ul><li>Delta Modulation (DM) and CVSD (continuously variable-slope DM) </li></ul></ul><ul><li>2. Audio Power Spectrum Coding. Examples include: </li></ul><ul><ul><li>Sub-band coding </li></ul></ul><ul><ul><li>RELP (regular - or residual - pulse excited linear predictive coding) </li></ul></ul><ul><ul><li>CELP (code excited linear predictive) </li></ul></ul><ul><ul><li>VSELP, ACELP (vector sum ELP, Algebraic code ELP) </li></ul></ul>
Some Speech Coder Bit-rates Typical Applications <ul><li>Lower bit rate coders are generally less satisfactory than higher bit rates. </li></ul><ul><li>*PCS= Personal Communications Service, a cellular radio system usually with digital speech coding </li></ul>
Non-voice “Bearer” Services <ul><li>Due to their near-ubiquitous presence, readily available investment capital, and the franchise held by many telephone operating companies to install wire, cable or fiber, many other services are also under development and use in the telephone system and related systems </li></ul><ul><li>Images: telefax, video, other images </li></ul><ul><li>Data: Internet access, data bases, and related information </li></ul><ul><li>Digital coding of any originally analog information (such as video) is seen as the optimum method for combined transmission </li></ul><ul><ul><li>but verify that the entire system is really advantageous!! </li></ul></ul>
Telephone Data Modems* <ul><li>Digital data can be transmitted via telephone voice channels using an audio frequency carrier signal which is modulated to convey binary information by changing its: </li></ul><ul><ul><li>Amplitude (instantaneous voltage or power level). This method is used alone only for Morse Code </li></ul></ul><ul><ul><li>Phase (relative time delay of oscillatory waveform peaks and valleys vis-à-vis a standard “clock” signal) </li></ul></ul><ul><ul><li>Frequency (the quantity of cycles per second; the musical pitch) </li></ul></ul><ul><li>Recent modem (modulator-demodulator) designs mostly use QAM (quadrature amplitude modulation) a combination of amplitude and phase modulation </li></ul><ul><li>V.90 or V.92: In one direction, various voltage amplitude levels are each used to represent a specific 7-bit binary data value. </li></ul><ul><li>ADSL: A special type of multi-carrier QAM modem is used via telephone subscriber wires to carry high bit-rate digital Internet signals in a frequency band above the usual voice frequencies. </li></ul><ul><li>* Modem is an invented word made of the first syllables taken from the two words Modulator and DEModulator. </li></ul>
Modem Properties <ul><li>Data modems today also include automatic equalizers to compensate for individual voice channel characteristics that would otherwise cause undesired waveform changes. </li></ul><ul><li>Data rates of up to 9.6, 14.4, 28.8 and 33.6 kb/s are feasible using classic adaptive QAM modem technology </li></ul><ul><li>Higher bit rates up to 56 kb/s* use direct PCM encoding at one end </li></ul><ul><ul><li>Fully digital connection at transmitting end. Analog connection at receiving end. Signal voltage can be measured with sufficient accuracy at receiving end to infer the PCM code value used. </li></ul></ul><ul><ul><li>Full 64 kb/s throughput requires a specifically installed digital line such as ISDN or DDS. </li></ul></ul><ul><li>* V.90 and V.92 modems today are legally limited to 53 kb/s. The highest voltage levels of PCM are prohibited to avoid “crosstalk” with other wire pairs in the same cables. </li></ul>
Fully Digital Telephone Services <ul><li>ISDN (integrated services digital network) and proprietary digital services (DDS, etc.) </li></ul><ul><li>Special digital signals used on the subscriber loop </li></ul><ul><li>Permits end-to-end 64 or 56 kb/s digital service </li></ul><ul><li>For voice, analog-digital conversion is performed in the ISDN telephone set rather than in the central office switch </li></ul><ul><li>Unfortunately ISDN is very costly, but has had a recent small surge in utilization due to Internet access applications. Some critics view ISDN as an early example of the “Iridium syndrome” </li></ul><ul><li>Emergence of 56 kb/s V.90/92 modems has severely reduced the use of ISDN </li></ul>
Packet Data Systems <ul><li>In several types of data networks, data is transmitted in “packet” format </li></ul><ul><ul><li>A small block of consecutive data bits from each particular source has a “header” pre-pended. The header contains, among other things, a code number indicating the destination. This is used to control routing. </li></ul></ul><ul><ul><li>Typically an error-detecting code is appended to the end of the packet. </li></ul></ul><ul><li>In some systems, all packets are the same “size” (length) – in others each packet is of different size, typically based on source data rate. </li></ul><ul><li>Packets from different sources are transmitted via the same channel, one after another </li></ul><ul><li>Most systems use a special “flag” bit pattern, 01111110, as a separator between packets. </li></ul><ul><ul><li>The internal packet bit stream is pre-modified (bit stuffing) to exclude any false occurrences of the”flag pattern. </li></ul></ul><ul><ul><li>At the receiving end, the bit stuffing process is undone </li></ul></ul>
Why Packets? <ul><li>Many types of digital information sources are “bursty” in time </li></ul><ul><ul><li>Brief “bursts”of high bit rate data are separated by some time intervals during which no data bits are generated </li></ul></ul><ul><ul><li>Data coding methods which remove redundant information from “raw” speech or video typically produce bursty data </li></ul></ul><ul><li>A number of different packet transmissions can be multiplexed on a shared channel in a high bit rate medium (co-ax, fiber, etc.) more efficiently than using a separate channel for each source, provided that all data sources do not continually produce data bursts simultaneously </li></ul>
ATM (Asynchronous Transfer Mode) <ul><li>ATM “Payload” data is transmitted in fixed size packets (called here “cells”) of 48 bytes (384 bits) with a 5 byte identification header (53 bytes total) </li></ul><ul><li>ATM signals can be transmitted e.g. via the “payload” of SONET/SDH at 50 Mb/s or more gross bit rate </li></ul><ul><li>Due to its small packet size, ATM has little signal delay, and is theoretically superior to other packet formats for digitally coded voice. </li></ul><ul><li>ATM is an interesting alternative to LAN/WAN technologies such as Ethernet, although presently far more costly </li></ul>
Telefax (Facsimile,FAX) <ul><li>Groups 1 and 2 FAX are obsolescent. </li></ul><ul><li>Group 3 FAX is in worldwide use. A page image is typically transmitted in less than a minute at 9.6 kb/s binary data rate via internal modem over a voice grade PSTN channel. </li></ul><ul><li>Group 3 FAX uses binary data compression coding of black/white pixel (or pel) picture elements (“dots”) </li></ul><ul><ul><li>Line difference coding takes advantage of vertical “lines” in the image </li></ul></ul><ul><ul><li>Run-length coding takes advantage of large contiguous areas of white or black, and of long run zero line differences produced by line difference coding. </li></ul></ul><ul><ul><li>Huffman coding takes advantage of repeated appearance of certain binary bit patterns in the FAX bit stream </li></ul></ul>
Other Data Compression Methods <ul><li>Lossless data compression methods exploit redundant data bit patterns when present, and accurately regenerate original data when decoded </li></ul><ul><ul><li>Plain language text has well-known frequently occurring characters (E T A O I N etc.) and infrequently occurring characters (J Z Q etc.), a fact that is exploited by Morse code and Huffman coding </li></ul></ul><ul><ul><li>LZW (Lempel-Ziv-Welch) coding dynamically adjusts transmission codes to use short binary patterns for frequent symbols and longer binary patterns for infrequent symbols. LZW is one type of “algebraic” coding. </li></ul></ul><ul><ul><li>Huffman coding is a non-dynamic formal lossless data compression method similar to LZW </li></ul></ul>
Lossy Coding <ul><li>The word “lossy” implies data compression with imperfect reconstruction of the original information </li></ul><ul><ul><li>Used when human perception can be “fooled,” for: </li></ul></ul><ul><ul><ul><li>Video and still pictures with continuous brightness range (gray or color) </li></ul></ul></ul><ul><ul><ul><li>Audio spectrum (speech) coding. Does not reproduce the exact sound waveform </li></ul></ul></ul><ul><li>Lossy image coding typically approximates the spatial brightness pattern using a family of orthogonal functions </li></ul><ul><ul><li>Discrete Cosine Transform is popular for images, video </li></ul></ul><ul><li>Frame difference and motion extrapolation methods are used with video as well </li></ul><ul><li>Video, which requires over 40 Mb/s with simple waveform coding, can be lossy-encoded at 64 to 128 kb/s (via Px64 code) with acceptable (not high) quality and coding delays </li></ul>
Error Protection Coding <ul><li>Use of additional bits with the “payload” data can be used to </li></ul><ul><ul><li>Correct a limited quantity of bit errors </li></ul></ul><ul><ul><li>Detect (but not correct) larger quantity of bit errors </li></ul></ul><ul><ul><ul><li>Error detection codes are often used in conjunction with an Automatic request to Retransmit (ARQ) strategy to retransmit pieces of the data (typically packets) unless they are soon acknowledged as received OK, via a message sent back to transmitter from the receiving end equipment. </li></ul></ul></ul><ul><ul><ul><li>Widely used with error-prone radio channels and delay-tolerant signals such as for wireless call processing messages </li></ul></ul></ul><ul><li>Also used in T-1 extended super frame version, and SONET/SDH multiplexing systems, to continuously monitor transmission accuracy. </li></ul>
Encryption <ul><li>Important when the transmission is physically open to interception by unauthorized persons </li></ul><ul><ul><li>Particularly for wireless, radio, microwave, etc. </li></ul></ul><ul><li>Encryption methods can also be used to authenticate messages </li></ul><ul><ul><li>Only a person who knows the correct “secret key” can properly encrypt or decrypt a message </li></ul></ul><ul><li>The most widely used physical level encryption method is the Vernam cipher </li></ul><ul><ul><li>A “secret” encryption cipher bit stream is “added” to the bits before transmission, then the same cipher bit stream is “subtracted” out at the receiver* </li></ul></ul><ul><ul><li>The practical complications in this process relate to generating, distributing and synchronizing the cipher bit streams at both the transmit and receive ends. </li></ul></ul><ul><ul><li>* Not normal arithmetic addition and subtraction. Rather the XOR, or ring sum or modulo-2 logical operation </li></ul></ul>
End of Lecture 1 <ul><li>The rest of the sessions involve a more detailed description of the technical topics just listed. </li></ul>