Paper 3 370

Telephone Almanac, American Telephone and Telegraph Company, 1922

In 1876, Alexander Graham Bell demonstrated the telephone at the Centennial Exposition of the United States in Philadelphia, Pennsylvania (United States). From that simple demonstration of one-way transmission over a distance of several hundred feet, the copper-based telephone network grew at an astounding rate. In 1880, Bell invented the first wireless communications system using reflected sunlight and photoelectric selenium receivers. Using this technique, however impractical, he was able to transmit intelligible speech a distance of up to 700 ft. Bell named this invention the photophone, later renaming it the radiophone, which he described as his greatest invention [1, 2]. AT&T continued work on the technology, extending its reach to several miles. The German Nazi military experimented with similar but more advanced systems in tank warfare during World War II [2].

The first commercial application of radio technology was that of broadcast radio, introduced in the United States in 1920. The first radio station, the Westinghouse station KDKA in Pittsburgh, Pennsylvania, inaugurated service by broadcasting returns of the Harding–Cox presidential election. On July 25, 1922, the first commercial station, WBAY, which was owned by AT&T, began broadcasting from the AT&T Long Lines building in New York City. Its first paying customer, two months later, was the Queensborough Corporation, advertising its Hawthorne Court real estate development in Jackson Heights. In the meantime, AT&T employees supplied the programming, which consisted of vocal selections, piano recitals, poetry recitations, and other content that seems fairly tame by today's standards. Among the performances was a recitation of James Whitcomb Riley's poem An Old Sweetheart of Mine by Miss Edna Cunningham; there is no record of the audience reaction [1]. While not an immediate commercial success, radio enjoyed U.S. market penetration of 50 percent of households in 1930, 90 percent in 1940, and essentially 100 percent by 1995.

Television, while first demonstrated in 1926 by John Logie Baier, a Scottish inventor, was refined by yet other Bell Labs employees, Frederick Eugene Ives and Frank Gray. The first public demonstration of color TV took place in the United States in 1927, between Bell Labs in Whippany, New Jersey, and Washington, D.C., where the audience included Secretary of Commerce Herbert Hoover. Vaudeville acts were included. Again, there is no record of the reaction of the audience. Delayed by the Great Depression and then World War II, commercial broadcast TV was introduced in 1946 and enjoyed success similar to that of radio [1]. Market penetration in the United States was estimated at 50 percent in 1955, 90 percent in 1960, and 99 percent in 1995. Wireless radio technology also was deployed early on in maritime communications, or ship-to-shore telephony and telegraphy. The first terrestrial mobile application was a one-way system employed in police radio dispatch trials in 1921 at the Detroit, Michigan, Police Department [2, 4].

11.1 WIRELESS DEFINED

11.4 CELL CONCEPT: FREQUENCY REUSE

Radio systems are designed for a certain area of coverage, or footprint. Even early radio and TV broadcast systems used the concept to define a service area, such as a metropolitan area. Therefore, you could reuse the same frequencies to support service in metropolitan areas some distance away. For example, 98.1 (MHz) on your FM dial might be WXYZ in New York, New York, and KFRC in San Francisco, California. Similarly, Channel 7 on your TV might be WFAA in Dallas, Texas, and KIRO in Seattle, Washington; broadcast TV stations in the United States can reuse frequencies if separated by at least 150 miles [5]. The size of the cells, of course, is sensitive to frequency and power level as well as the height of the antennas, the topography, the time of day or night, solar activity, weather conditions, and other factors.

The formal concept of radio cells dates back to 1947, when Bell Telephone engineers developed a radio system concept that included numerous, low-power transmit/receive antennas [1]. Scattered throughout a metropolitan area, this sort of architecture served to increase the effective subscriber capacity of radio systems by breaking the area of coverage into cells, or smaller areas of coverage. Thereby, each frequency could be reused in nonadjacent cells. Additionally, the cells can be split, or subdivided, further as the traffic demands of the system increase. In other words, cellular radio networks are highly scalable.

Cells can be characterized as falling into three broad descriptive categories: macro-cells, microcells, and picocells, as illustrated in Figure 11.1. As the cells shrink, the advantages of frequency reuse increase significantly. The increase in traffic-handling capacity can be remarkable, with associated increases in revenue potential. However, the costs of network deployment increase along with the number of base stations and interconnecting links. Also, the complexity of switching mobile traffic between cells increases considerably.

Assume that 12 channels are available for use in a metropolitan area that is 60 miles wide (i.e., has a diameter of 60 miles and a radius of 30 miles) and that the terrain is perfectly flat. Assume a seven-cell reuse pattern, which is typical. Consider the following theoretical scenario:

· Picocells are quite small, covering only a few blocks of an urban area or, perhaps, a tunnel, walkway, or parking garage. In a seven-cell reuse pattern, with each cell covering a radius of approximately one-half mile, the reuse factor climbs to 514. In other words, the same 12 channels could theoretically support up to 6168 simultaneous conversations [2].

Within a given cell, the antenna may operate in either an omnidirectional or a vectored mode. The omnidirectional approach involves signal transmission from and signal receipt by the centralized base station antenna in all directions. This approach uses all frequency channels in a 360° beam, and all at the same power level. A vectored antenna can subdivide the coverage area of the cell into multiple vectors of coverage. Cellular antennas, for example, commonly carve a cell into three vectors, each of 120°. This vectoring approach enables the carrier frequencies within each vector to be managed independently from those within the other vectors in terms of dimensions such as channel allocation and signal strength.

Smart antennas can improve significantly on this concept of vectoring. Such antennas can operate within a given cell and its constituent vectors, further subdividing the coverage through the use of as many as 12 beams, each with a beam width of 30 °. Thereby, the footprint of the cell can be sculpted to optimize coverage, channel allocation, and interference. The individual beams can rotate from one vector to another and can do so on a dynamic basis in order to adjust to traffic patterns that might vary by time of day, with the process of beam management performed remotely [6].

At the most basic level, frequency division is the starting point for all wireless communications because all communications within a given cell must be separated by frequency to avoid their mutual interference. Frequency Division Multiple Access (FDMA) divides the assigned frequency range into multiple frequency channels to support multiple conversations, as depicted in Figure 11.3. In other words, multiple narrowband frequency channels derive from a wider band of assigned radio spectrum, much as frequency division multiplexers (FDMs) operate in the wired world. Multiple incoming and outgoing calls contend for access to those channels. A given call takes place on one pair of frequencies, with one transmission in the forward direction and another for transmission in the reverse direction. At the same time, another call takes place on another pair of frequencies. The forward and reverse channels in each frequency pair are separated in frequency in order to avoid cochan-nel interference. The forward channels are separated from each other, as are the reverse channels, in order to minimize the potential for crosstalk. The station equipment must be frequency agile in order to search for and seize an available frequency channel, especially as the mobile transmitter/receiver moves from one cell to another in a cellular network.

Frequency Division Duplex (FDD) is a means of providing duplex (bidirectional) communications. Forward and backward channels make use of separate frequencies. FDD is used with both analog and digital wireless technologies, including cordless telephony and cellular.

Time Division Multiple Access (TDMA) is a digital technique that divides each frequency channel into multiple time slots, each of which supports an individual conversation (see Figure 11.4). This concept is exactly the same as in the wired world, where time division multiplexers perform the same function in a T/E-carrier environment. The total available bandwidth, the bandwidth of the individual channels, and the number of time slots per channel vary according to the particular standard in place, as well as the specific coding technique employed. GSM, for example, involves a carrier channel of 200 kHz, with a channel rate of approximately 200 kbps. The channel is divided into eight time slots of 25 kbps each, easily supporting low-bit-rate digitized voice of 9.6 kbps, plus overhead for framing and signaling. Each set of eight time slots is organized into a logical frame, and the frames are repeated frequently. Each conversation makes use of two time slots, one for the forward channel and one for the reverse channel. Assuming the same level of bandwidth, systems based on TDMA offer roughly three to four times the traffic capacity of those based on FDMA [1]. TDMA was specified first in EIA/TIA Interim Standard 54 (IS-54) and later was included in IS-136, which is an evolved version of IS-54 for use in cellular and PCS systems. TDMA is used in systems based on the D-AMPS and GSM cellular standards, and DECT cordless telephone standards.

E-TDMA (Enhanced TDMA), developed by Hughes Network Systems, is an improvement over TDMA, employing Digital Speech Interpolation (DSI) compression, also known as Voice Activity Detection (VAD), and half-rate vocoders (voice coders) operating at 4.8 kbps to enhance bandwidth utilization. E-TDMA provides as much as 16: 1 improvement over analog technology.

Time Division Duplex (TDD) is a digital means of providing bidirectional communications. TDD can be employed with both channels using the same frequency, but this ping-pong transmission approach can yield poor quality, as it actually is a half-duplex (HDX) transmission mode. (Note: In a properly engineered network, this HDX mode can be made transparent.) More commonly, TDD is used in conjunction with FDD, with the forward and backward TDM channels riding over separate frequency channels. Further, the time slots are staggered so the frequency-specific transceivers are not asked to transmit and receive at the same exact points in time. Time slot 1, for example, might be used on the forward channel, or uplink, from the user terminal to the centralized cell antenna, while time slot 3 is used on the backward channel, or downlink, from the cell site to the user terminal.

In addition to its enhanced traffic capacity when compared to FDMA, TDMA supports data traffic as well as voice. TDMA also offers considerable flexibility, as multiple time slots can be assigned to an application, depending on its bandwidth requirements.

Code Division Multiple Access (CDMA) is a relatively new technology that has its roots in Spread-Spectrum (SS) radio. Hedy Lamarr, the famous actress and dancer of pre–World War II fame, created the concept of spread spectrum in 1940. As the story goes, Lamarr developed spread spectrum radio in order to remotely synchronize multiple player pianos. Droves of people supposedly paid good money to go to the resulting radio-controlled piano concerts in this much simpler time. The U.S. Patent and Trademark Office issued a patent to Ms. Lamarr and George Antheil, a film-score composer to whom she had turned for help in perfecting the idea, for a secret communication system that was, in effect, a spread spectrum radio. In the Pacific Theater during World War II, the Allies used that patented technology extensively to prevent the Japanese from jamming radio-controlled torpedoes. This primitive system used a mechanical switching system much like a piano roll to shift frequencies faster than the Nazis or the Japanese could follow them. Subsequently, spread spectrum has combined with digital technology for spy-proof and noise-resistant battlefield communications. During the 1962 Cuban nuclear missile crisis, for example, Sylvania installed it on U.S. warships sent to blockade Cuba, where the technology provided improved security as well as prevented signal jamming. Ms. Lamarr never asked for, and never received, any royalties from the use of her invention. Ms. Lamarr was quite an innovator, by the way, and across multiple disciplines. She delighted and shocked audiences in the 1930s by dancing in the nude in the movie Ecstacy [7].

· Direct-Sequence Spread Spectrum (DSSS) is a radio technique in which the narrowband signal is spread across a wider carrier frequency band. Each transmission is assigned a 10-bit pseudorandom binary code sequence, which comprises a series of 1s and 0s in a seemingly random pattern known to both the transmitter and receiver. The original code sequence is mathematically self-correlated to yield a code that stands out from all others, at least on average. The paired transmitters and receivers recognize their assigned and correlated code sequences, which look to all others as PseudoNoise (PN). DSSS phase modulates the carrier wave with a continuous string of PN code symbols, or chips, each of which has a much shorter duration than a data bit. So, the chip rate is much faster than the bit rate. Thereby, the noise signal occurs with a much higher frequency than the original data signal and spreads the signal energy over a much wider band (Figure 11.5). Rake receivers are used in CDMA systems to improve signal quality.

CDMA improves bandwidth utilization because a great number of users can share the same wideband radio frequency channel (see Figure 11.5). While initial predictions were that CDMA would improve on AMPS by a factor of up to 20: 1, practical results have been more in the range of 15: 1 in cellular telephony applications. Particularly through the use of the FHSS approach, CDMA also provides excellent security, as it is virtually impossible to intercept more than a small portion of a transmission. Encryption, of course, can provide additional security [8]. FHSS also offers the advantage of improved overall transmission, since no individual transmission gets stuck with the assignment of a poor-quality channel.

Qualcomm perfected and commercialized CDMA and has gone on to develop, manufacture, market, and license CDMA products. The first commercial CDMA system was placed in service in Hong Kong, where cellular phones have long been considered a necessity and where the networks have suffered terrible congestion. Since then, a great number of manufacturers and providers of cellular, PCS, WLANs, and other systems and networks have licensed CDMA.

The best commonsense means of comparing FDMA, TDMA, and CDMA is to use the cocktail party analogy, which has been used by acoustical engineers for many years. This is an international cocktail party with a great number of people all wanting to talk at the same time. The cocktail party is being held in a large ballroom (frequency band).

In an FDMA environment, each conversation takes place in its own space (channel). Therefore, the ballroom must be subdivided into smaller rooms (cells) of suitable size (bandwidth) and then subdivided into smaller cubicles (channels) so there is no interference between the pairs of people engaged in each conversation. At any given time during the conversation, each pair occupies a separate room. The pairs can move from room to room (cell to cell) as they make their way from the entrance to the buffet, although they must interrupt their conversations when they leave one room and must enter another empty room before resuming conversation (hard handoff, or break and make). The number of simultaneous conversations that can be supported depends on the number of rooms of suitable size that can be derived from the space available in the ballroom. The speakers also must control their volume and frequency levels in consideration of the thickness of the walls (guard bands, or frequency separations) so they do not interfere with others.

In a CDMA environment, the entire ballroom is open. A larger number of people can fit into the room because the walls have been removed. A larger number of pairs of people can engage in simultaneous conversations, with the conversations overlapping in frequency, volume, and time. Volume and frequency levels still must be controlled, of course. As highly intelligent receivers, the listening parties can distinguish the transmission of their paired speakers by locking on the language (PN code) spoken, even given the high level of background noise (interference) from other conversations. Since the walls have been removed, the pairs can move from one area to another without losing conversational connectivity (soft handoff, or make and break)

11.6 SPECIALIZED MOBILE RADIO

Two-way mobile radio dates to the very early days of radio. The U.S. Army Signal Corps mounted early spark transmitters in vehicles in 1904 and experimented with air-to-ground communications in 1908. The Detroit, Michigan, Police Department placed the first experimental one-way mobile radio dispatch system into service in 1921, operating in the 2-MHz band with the call letters KOP [9]. The first two-way mobile system was installed by the Bayonne, New Jersey, Police Department in the early 1930s. While this Amplitude Modulation (AM) radio application grew quickly, even as late as 1937 the FCC allocated only 40 channels to mobile radio. In the late 1930s, Frequency Modulation (FM) replaced AM as the method of choice because of its improved quality of reception and lower power requirements. (FM receivers tend to lock in on the stronger competing signal, whereas AM recognizes all competing signals.) In 1949, the FCC recognized two-way mobile radio as a new class of service and began to allocate more spectrum and to regulate its use [2, 10].

In 1946, AT&T was granted the first commercial license for two-way, mobile FM service. That first system in St. Louis, Missouri (United States), employed a centralized antenna with a range of 50 miles. The system not only served to interconnect mobile phones but also provided connection to the Public Switched Telephone Network (PSTN). As the service was fairly inexpensive and extremely convenient, it quickly grew in popularity and soon was oversubscribed. Similar systems were quickly deployed in other major U.S. cities, where they, too, were soon oversubscribed. In fact, it was not uncommon for a provider to load as many as 100 subscribers per channel, which resulted in horrible service quality. In 1976, for example, service in the New York metropolitan area consisted of 20 channels supporting 543 subscribers out of a total population of approximately 20 million. Not surprisingly, there was a waiting list of approximately 3700 [10].

SMR/TMR largely has been supplanted by cellular service offerings, although it remains widely used in dispatch and fleet applications such as police, fire, and emergency vehicles as well as taxi fleets, utility fleets (e.g., telephone companies, gas and electric utilities, and CATV providers), and courier services. In the United States, 80 MHz has been allocated for SMR.

The walkie talkie was invented in 1938 by Al Gross, a high school student in Cleveland, Ohio (United States), at the time. The portable hand-held radio transmitter–receiver caught the attention of the Office of Strategic Services (OSS), predecessor to the Central Intelligence Agency (CIA). The OSS recruited Gross, who then led the effort to develop the walkie talkie for clandestine and military uses. Code named Joan/Eleanor, the first walkie talkie system comprised a ground unit, Joan, and an airborne unit, Eleanor. The system allowed OSS agents behind enemy lines to communicate with aircraft in a manner that virtually defied detection at the time. Gross also is credited with inventing the pager, the CB radio, and the cordless telephone. Gross also lobbied the FCC to create the Personal Radio license spectrum, which later became Citizens Radio Service Frequency Band, or CB radio. Gross formed the Citizens Radio Corporation, which manufactured and sold personal two-way radios, mostly to farmers and the U.S. Coast Guard [11].

11.7 PAGING

The paging system was invented by Al Gross as an adaptation of his two-way radio, the walkie talkie. After some early market resistance from doctors who were afraid that the system would upset their patients and disturb their golf games, Gross sold the first system in 1950 to New York's Jewish Hospital [11]. That first system provided a means by which a centralized antenna could broadcast alerts to small, inexpensive pagers, or beepers. A page simply transmitted an identification number, which was recognized only by the pager being addressed. If that pager were in range, it beeped, hence the term beeper. Response to the page was in the form of a telephone call to the paging company to retrieve a message. The FCC approved pagers for consumer use in 1958. The first consumer pager was the Motorola Pageboy I, which was based on the proprietary protocols including the GOLAY standard [10].

During the 1970s, an international team of radio engineers developed a standard set of code and signaling formats. That effort evolved into the POCSAG (Post Office Code Standardization Advisory Group) code, the name of which was derived from the fact that the British Post Office (BPO), which was the PTT for the United Kingdom at the time, chaired the effort. The POCSAG standard, which is in the public domain, provides for transmission speed of up to 2400 bps using channels of 25 kHz in the band 150–170 MHz. The CCIR (now ITU-R) standardized that code internationally in 1981, and most nations quickly adopted it. POCSAG can support as many as 2 million individual pager addresses. Tone-only, numeric, and alphanumeric pagers are supported on a one-way basis.

Motorola more recently floated the FLEX set of proprietary solutions, which largely have replaced POCSAG in the United States and have become de facto standards throughout most of the world, excepting Western Europe. Those solutions provide two-way messaging, support data transmission, and provide greater bandwidth. FLEX also supports as many as 5 billion addresses, with up to 600,000 supported per channel. The FLEX family of protocols includes the following [12–15]:

· FLEX: 1600bps; 25-kHz channels; one-way

· ReFLEX: 1600, 3200, 6400, or 9600 bps; 25- or 50-kHz channels downstream and 12.5-kHz channel upstream; two-way

· InFLEXion: Up to 112 kbps; 50-kHz channels in the N-PCS (Narrowband PCS range); two-way; supports compressed voice downstream

Pagers generally operate over 25-kHz channels in the 900-MHz band. In 1984, the FCC (U.S.) dedicated 1 MHz of 40 channels in this band for nationwide paging purposes. RCCs (Radio Common Carriers) and PPOs (Private Paging Operators) provide paging services. Regulated by the FCC and the state Public Utility Commissions (PUCs), RCCs make use of FCC-designated frequencies. PPOs are unregulated but must share spectrum with other users in the VHF and UHF bands.

A typical page begins with a message transmitted to a centralized Network Operations Center (NOC). Using SkyTel, a Verizon company, as an example, the NOC forwards the page to a satellite with an appropriate footprint, or coverage area. As depicted in Figure 11.7, the satellite forwards the page to a terrestrial network of centralized antennas interconnected through various means, including private microwave, leased lines, and Frame Relay. The terrestrial antenna network forwards the page to the pager. In a two-way paging scenario, the response travels back to the terrestrial base station and then over a Frame Relay network to the NOC, where it is forwarded to the party who originated the page.

Generally speaking, these satellite links are highly beneficial and highly reliable. The May 1998 failure of the Galaxy IV satellite, however, underscored the vulnerability of even the most sophisticated networks. When the satellite lost its orientation, hundreds of terrestrial paging antennas had to be reprogrammed by hand in a process that took over a week. In addition to the 40 million or so pagers that lost service during this outage, 5400 of 7700 Chevron gas stations lost their pay-at-the-pump capabilities, and music-on-hold systems went down, causing callers to hang up when they thought they had been disconnected. Additionally, the results of the multistate Powerball lottery drawing could not be broadcast to the 88 TV stations that regularly carry that programming [16].

Two-Way Paging (TWP) systems have been available abroad for years but were not introduced in the United States until roughly 1995. The simplest and most common version is known as 1.5-way paging. This approach supports guaranteed message delivery, as the network does not attempt to download messages until such time as the pager is within range, is turned on, and has enough memory to support the download. The general location of the pager is communicated upstream, so the messages can be downloaded to the antennas supporting that particular geographic area, rather than being broadcast across the entire paging network. Once downloaded successfully, the pager acknowledges to the network the receipt of the page. Such pagers are alphanumeric, supporting the display of a text message and return telephone number, typically for a telephone or pager. Full two-way paging enables the recipient of the page to select and transmit a return message, which most commonly is selected from a small set of predetermined messages, although some pagers support user-definable return messages.

Pagers also have been incorporated into other devices, such as cellular telephones, Personal Digital Assistants (PDAs), watches, and even key chains. Whether standalone or merged into another device, contemporary pager capabilities include some combination of alphanumeric display, two-way communications, message storage, audible and vibration alert, fax receipt (very unusual), and abbreviated e-mail forwarding (very unusual). Many paging service providers also offer information such as weather reports, sports scores, traffic reports, and stock quotes, all of which are provided through a feature known as Short Message Service (SMS), which currently is limited to 160 characters [17]. Limited Web content also may be available. Incoming messages generally may be reviewed, erased, and archived. Both incoming and outgoing messages may be date and timestamped.

Pagers reached peak popularity in the 1980s and 1990s. In 1994, there were estimated to be well over 61 million pager subscribers worldwide, according to some estimates [18]. In the United States alone, there were over 27 million subscribers to over 2000 paging services [19]. At that point, pagers were overwhelmed by cellular telephones, which offer much more functionality, although at considerably higher cost. As a result, millions of pagers have been disconnected over the past decade or so. Pagers remain used in areas where cellular service is not available and in applications where cellular functionality is undesirable from a cost standpoint.

Paging costs generally involve a flat monthly fee which includes some number of pages. The cost of pages above the threshold generally is sensitive to their nature (e.g., numeric, alphanumeric, or voice). A few paging service providers offer the option of paging party pays, which charges the paging party, rather than the paged party, for the page.

11.8 CORDLESS TELEPHONY AND WIRELESS OFFICE TELECOMMUNICATIONS SYSTEMS

Generally an extension and application of common cordless telephony, WOTS provides the advantage of mobility for a small group of select employees who must have the freedom to wander around the complex but must have communications capability at all times. However, WOTS PBX adjuncts tend to be highly capacity limited and costs can be easily double those of traditional phones when the additional infrastructure costs are taken into consideration. WOTS falls into the general category of low-tier systems, which are intended for pedestrian, in-building, and Wireless Local Loop (WLL) application. There exist multiple cordless telephony standards used to provide WOTS service. The original cordless telephones (c. 1980) in the United States were assigned 1 of 10 channels in the 27-MHz range. As these early analog phones transmitted in the clear, that is, with no encryption or signal-scrambling mechanism, and as there were so few channels, it was not unusual to accidentally engage in a three-way conversation with a neighbor who happened to have a phone operating on the same channel. In 1986, the FCC changed the cordless frequency range to the 46-and 49-MHz bands and reduced the allowable power levels, which had the effect of reducing the range slightly but improving signal quality. Contemporary digital versions operate in the 900-MHz and 2.4-GHz bands, which are in the unlicensed Industrial/Scientific/Medical (ISM) band. Contemporary standards include the following, the most significant of which are compared in Table 11.1:

· CT1 (Cordless Telephony generation 1) was developed in Europe, where it was known as CEPT-1. CT1 operates in the 915-and 960-MHz bands over 40 paired channels 25 kHz wide. FDMA and FDD derive two separate channels, one for transmission and one for reception, each of which is 12.5 kHz wide. CT1 is analog and low cost but limited in range to 150m or so. FM was employed in this voice-only technology. A variation on this standard is CT0, which was primarily used in the United Kingdom. CT0 specified eight paired channels, with the base station transmission in the 1.642–1.782-GHz range and portable station transmission in the 47-MHz range. A number of parochial CT0 versions were developed in other countries.

· CT1 + is a variation on CT1 developed in concert by Belgium, Germany, and Switzerland. CT1+ was intended as the basis for a public wireless service, along the lines of Telepoint. CT1+ operated in the 887-and 932-MHz bands over 80 channels 25 kHz wide. FDMA and FDD derive two separate channels, one for transmission and one for reception, each of which is 12.5 kHz wide. Although CT1+ was not successful, it did originate the concept of a Common Air Interface (CAI), which enables multiple manufacturers to develop products in support of a public cordless telephony service offering.

· CT2 + is an improvement on CT2, supporting two-way calling and call hand-off. CT2+ uses 8 MHz of bandwidth in the 900-MHz range. CT2+ is based on dynamic channel allocation, requiring frequency-agile handsets. Encryption is supported for improved security. A common signaling and control channel offers improved call setup times, increased traffic capacity, and longer battery life because the handset must monitor only the signaling channel. CT2+ has been used in applications such as the Walkabout public cordless telephony trial in Canberra, Australia.

· CT3 was developed by Ericsson in 1990 as a proprietary solution designed for high-density office environments. CT3 is based on TDMA and TDD and uses the same frequency bands as CT2+, supporting roaming and seamless hand-off. CT3 supports Telepoint (PCS) and in-building office applications (WOTS).

· Personal Handyphone System (PHS) is a digital cordless telephony system developed in Japan and used in China, Taiwan, and throughout Asia. PHS carves out 77 carriers with spacing of 300 kHz in the band 1895–1918.1 MHz. The upper half of the band is used for public systems and the lower half for home office applications. Each carrier supports four TDMA channels employing π/4 Differential Quaternary Phase Shift Keying (π /4 DQPSK) to yield a theoretical data rate of 384 kbps. Voice is encoded at 32 kbps. PHS enjoyed considerable market success but lately has declined in the face of competition from low-cost cellular offerings. (Note: π is the Greek letter pi, which is the symbol for the ratio of the circumference of a circle divided by its diameter. Pi is a mathematical constant with an approximate value of 3.14159. Pi is from the Greek periphereia, which translates into periphery, i.e., circumference.)

· Personal Communications Services (PCS) in the United States is based on the Wireless Access Communication System (WACS), which subsequently was modified to an industry standard known as Personal Access Communications Services (PACS). PCS employs FDD to carve out 16 paired downstream and upstream carriers with spacing of 300 kHz in the bands 1850–1910 and 1930–1990 MHz. Within each frequency channel, PCS derives eight TDMA channels employing π /4 Quadrature Phase Shift Keying (π /4 QPSK). Voice encoding is ADPCM at 32 kbps and the theoretical data rate is 384 kbps. PCS has never gained any traction in the United States due largely to the prevalence and low cost of cellular telephony [23, 24, 26, 27].

This original concept is the basis of cellular radio. The first prototype system was developed by AT&T and Bell Labs in 1977. In 1979 in Tokyo, Nippon Telephone and Telegraph (NTT) activated the first commercial cellular system, utilizing 600 duplex analog radio channels in the band 925–940 MHz for the uplink channels from the Mobile Station (MS) to the Base Station (BS) and the band 870–885 MHz for the downlink, or reverse, channels [26]. The first system in the United States was activated in Chicago on October 13, 1983. AT&T operated that AMPS network for exactly 79 days, at which point the Modified Final Judgment (MFJ) took effect and a subsidiary of Ameritech (now part of SBC) assumed ownership [2]. By 1984, the Chicago network already was saturated in some cells and cellular telephones were in the hands of 91,600 people, growing to 19 million by 1994 [2, 3]. A cellular telephony network comprises multiple low-power transmit/receive antennas distributed throughout a geographic area, with each cell site having a relatively small, more or less circular area of coverage (see Figure 11.9). The coverage area of each individual cell overlaps those of neighboring cells, with the cell diameter generally a minimum of about 1 mile and a maximum of about 5 miles, sensitive to factors such as topography and traffic density [8]. As the terminal device moves out of the effective range of one cell, the call switches from one cell antenna to another through a process known as hand-off in order to maintain connectivity at acceptable signal strength. The hand-off is controlled through a Mobile Telephone Switching Office (MTSO), which is the functional equivalent of the PSTN Central Office (CO). The MTSOs generally are interconnected and connected to the PSTN through either private microwave or leased-line facilities.

Each cell site supports a limited number of frequency channels in order to take advantage of frequency reuse. The U.S. analog AMPS networks, for example, divide 333 frequencies (416 in certain areas) per carrier among cell sites; the average cell site supports 56 channels. To improve the performance of the network by reducing crosstalk, the original omnidirectional antennas were replaced with vectored versions. Vectoring generally involves three vectors, each with coverage of 120 °. Splitting the frequencies into three vectors had the unfortunate effect of reducing by two-thirds the number of channels available to any given user in the cell site. As a result, more cell sites of smaller size were required, which entailed additional cost. For aesthetic reasons, however, many cities and towns placed moratoria on the construction of new antenna sites. In response, antennas were developed that can support the full 333/416 channels, subdividing the area of coverage through the use of 12 beams, each with a 360 ° sweep. These smart antennas communicate continuously with the MTSO so that channel allocation is managed cell by cell and on the basis of the network as a whole and in consideration of shifts in traffic patterns [16].

Cellular standards are numerous and largely incompatible. Standards include both early analog and more recent digital solutions. Digital systems offer the advantages of improved error performance, improved bandwidth utilization through compression, and enhanced security through encryption and other mechanisms. Digital systems also support data communications much more effectively. In the United States, carriers are shifting subscribers to digital systems through various marketing enticements. The FCC has authorized U.S. carriers to cease support for analog cellular systems as of March 1, 2008.

The first cellular systems were analog in nature. Collectively, these analog solutions are categorized as 1G (1st Generation) systems. They include the following and are compared side by side in Table 11.2.

Also developed by Motorola, N-AMPS enhances the performance of an analog AMPS system. System capacity is improved by splitting each 30-kHz channel into three 10-kHz channels, thereby tripling AMPS capacity. Very few U.S. carriers deployed N-AMPS.

TACS is a derivative of AMPS developed for use in the United Kingdom in the 900-MHz band. TACS supports either 600 or 1000 channels, each of 25 kHz, compared with the 666/832 channels supported by AMPS. A number of variations were developed, including Narrowband TACS (NTACS), Extended TACS (ETACS), and Japanese Total Access Communications System (JTACS). TACS found acceptance in very few nations, largely has been replaced by GSM, and is considered obsolete in the United Kingdom.

Digital cellular clearly dominates the cellular radio world, having almost completely replaced the analog systems. There are a great number of standards, and a number of standards-based solutions, none of which are compatible. Collectively, these systems are grouped into the 2G (2nd Generation) category. They include the following, the most important of which are compared side by side in Table 11.3.

PCS is a U.S. term for cellular systems based on EIA/TIA IS-95a, also known as cdmaOne and CDMA Digital Cellular. The first commercial systems were installed in South Korea and Hong Kong in 1995 and in the United States in 1996. PCS uses one or more frequency bands of 1.25-MHz converted from the existing AMPS spectrum of each carrier deploying the service. Each 1.25-MHz band is subdivided into 20 carriers, each of which can support as many as 798 simultaneous calls and aggregate bandwidth of up to 1.288 Mbps. Because CDMA is employed, no guard bands are required. IS-95 can support dual-mode communications, operating in the same network as AMPS. As previously noted, CDMA offers the advantages of improved bandwidth utilization as compared to AMPS (as much as 10: 1 or even 20: 1) and TDMA (as much as 6: 1), soft hand-off, variable-rate speech-encoding, and support for both voice and data. The basic user channel rate is 9.6 kbps, although various channel rates can be achieved depending on the carrier implementation. The variable-rate speech-encoding algorithm runs at maximum rates of 8 kbps using Code-Excited Linear Prediction (CELP) or 13 kbps using Enhanced Variable-Rate voCoder (EVRC) and varies the rate downward to as low as one-eighth rate if the level of speech activity permits. The IS-95-B specification supports symmetric data rates of 4.8 and 14.4 kbps per channel; as many as eight channels can be aggregated to support a data rate up to 115.2 kbps. CDMA offers additional advantages in terms of maximum cell size due to improved antennas sensitivity and battery time due to precise power control mechanisms. IS-95 has been deployed by Airtouch (now part of Verizon), Ameritech Cellular (now part of Cingular), AT&T Wireless (now part of Cingular), Bell Atlantic Mobile (now part of Verizon), GTE MobilNet (now part of Verizon), 360 ° Communications (now part of ALLTEL), and Sprint Nextel [28–31].

Previously known as Japanese Digital Cellular (JDC), PDC is a Japanese standard for a digital system operating in the 800-, 900-MHz, and 1400-MHz frequency ranges. PDC derives 1600 RF carriers, each of which supports three TDMA channels at 42 kbps. Full-rate voice encoding is at 9.6 kbps and half-rate at 5.6 kbps. PDC has not found acceptance outside Japan, where it was extremely popular, with the exception of a few Asian countries under Japanese economic influence. PDC is being phased out in Japan, in favor of cdma2000.

The 1G analog networks really were not designed with data communications in mind. Over time, a number of manufacturers developed modems that could achieve theoretical data rates of as much as 28.8 kbps under optimum conditions, but conditions were rarely optimum. Factors affecting signal quality include the distance between the mobile station and the base station and Multi-Path Interference (MPI). Given the high error rates typical in cellular networks, the modem protocol either must include Forward Error Correction (FEC), which is overhead intensive, or must provide for repeated transmissions of errored and lost data packets or both. All things considered, throughput generally was more in the range of half the advertised rate at best. As analog systems take a break-and-make approach to call hand-off between cells, dropped connections were routine for the mobile cellular data user.

The 2G digital cellular systems improved on the analog approach, offering all digital data communications advantages enjoyed in the wired world. Those advantages include more usable bandwidth, improved error performance, and therefore enhanced throughput. Issues of inherent error performance and MPI remain, however. The hard hand-offs associated with the break-and-make connection technique persisted in 2G networks, with the exception of GSM, which used the softer approach of make-and-break to maintain connectivity.

In support of data communications over cellular networks, several manufacturers unveiled data-ready cellular phones as early as January 1995. Display screens varied in size from 16 to 160 characters. In addition to voice and SMS, those early data-ready phones supported Internet access, facsimile, and e-mail [33]. Nonetheless, the early cellular networks were not particularly friendly toward data communications.

Beginning in 1992, the ITU-R began work on International Mobile Telecommunications-2000 (IMT-2000). Variously dubbed 2.5G (i.e., the generation midway between 2G and 3G) and 3G, some of the many standards are well developed and widely available while others require a bit more work. They all grew out of an attempt to define a single global standard, which effort collapsed in October 1999, when representatives from the various countries agreed to adopt federal standards under the IMT-2000 umbrella. Two of the three modes are based on Code Division Multiple Access (CDMA) and one on Time Division Multiple Access (TDMA). cdmaOne, also known as Telecommunications Industry Association Interim Standard-95a (TIA IS-95a), was the first CDMA-based 2G approach to be introduced and is popular with CDMA-based cellular operators in North America and Asia. CDMA2000 3XMC, the high-speed version operating at 2 Mbps, was developed by Qualcomm and has been approved by the ITU-R. Enhanced Data Service for GSM Evolution (EDGE) is the TDMA variant of IMT-2000. General Packet Radio Service (GPRS), an interim step toward EDGE, is the choice of cellular operators with networks based on Global System for Mobile Communications (GSM) and Digital Advanced Mobile Phone Service (D-AMPS). Regardless of the transition approach and the specifics of the technology, all 3G systems will be based on CDMA.

In general, these standards support much increased data rates for both mobile and fixed wireless versions. In general, each has its roots in and is an extension of a 2G technology. So, the incompatibilities roll on, although at faster speeds, in support of voice, Short Message Service (SMS) and Multimedia Messaging Service (MMS), and Internet access. The 2.5G networks run over the same spectrum as the 2G networks and are backward compatible with them. Hence the transmission rates are lower than those of 3G networks, which run the over the spectrum set aside for their specific purposes. 2.5G and 3G technologies and standards include EDGE, GPRS, IMT-2000, UMTS, and W-CDMA. All of these approaches fit under the umbrella of International Mobile Telecommunications-2000 (IMT-2000), which replaced Future Public Land Mobile Telecommunications System (FPLMTS) as the vision for a single global standard for wireless networks. IMT-2000 is an ITU initiative for a twenty-first century wireless network architecture. As illustrated in Figure 11.10 and compared in Table 11.4, specifications include 128/144 kbps for high-mobility applications, 384 kbps for pedestrian speed (i.e., walking speed) applications, and 2.048 Mbps for both fixed WLL (Wireless Local Loop) and in-building applications such as WLANs (Wireless LANs). IMT-2000 also was intended to operate in the 2-GHz band. Note that 2000 has several meanings: the year 2000, bandwidth of up to 2000 kbps, and frequency range 2000 MHz. As it turned out, the year 2000 was pushed back to at least 2002, bandwidth of 2 Mbps is only for limited in-building applications, and the 2-GHz spectrum has been allocated for other pur-poses in North America as part of the unlicensed ISM band.

General Packet Radio Service (GPRS) is the 2.5G data service enhancement for GSM host networks. GPRS specifications were developed in 1997 by ETSI, which has since passed that responsibility on to the 3rd Generation Partnership Project (3GPP). GPRS is a packet-switched service that takes advantage of available GSM time slots for data communications and supports both X.25 and TCP/IP packet protocols with QoS. GPRS, an important component in the GSM evolution, enables high-speed mobile datacom usage and is considered most useful for bursty data applications such as mobile Internet browsing, e-mail, and various push technologies. Through linking together as many as eight GSM channels, GPRS has a theoretical transmission rate as high as 171.2 kbps, although it realistically is limited to 115.2 kbps and more typically 30–60 kbps. In practice, however, GSM system operators are unlikely to allow a single user to access eight channels. GPRS maintains the same GMSK modulation scheme used by GSM and provides always-on access, although charges apply only for actual data traffic.

Notably, GPRS will support simultaneous voice and data communications over the same wireless link, with voice taking precedence as always. GPRS defines three classes of terminal equipment:

· Class A terminals support simultaneous circuit-switched GSM voice and SMS service as well as GPRS packet-switched data traffic.

· Class B terminals will support nonsimultaneous circuit-switched voice and packet-switched data, automatically switching between the two. A Class B ter-minal, for example, will suspend an active data session in the event of an incoming voice call or SMS message. Most GPRS terminals are Class B.

· Class C terminals support either circuit-switched voice and SMS service or packet-switched services but must be manually switched from one to the other.

The bit rate is sensitive to the encoding scheme in use, of which there are four. CS-4 supports a bit rate of 20.0 kbps per time slot but can be used only when the Mobile Station (MS) and Base Station (BS) are in proximity. As the distance increases between the MS and BS, the encoding scheme must be more robust to compensate for attenuation and MPI, and the bit rate accordingly must adjust downward. At the edge of the cell, for example, CS-1 supports a bit rate of only 8.0kbps. The four Coding Schemes (CSs) and associated bit rates are as follows:

· CS-4: 20.0kbps

· CS-3: 12.0kbps

· CS-2: 14.4kbps

· CS-1: 8.0kbps

The U.S. carriers deploying GPRS include Cingular (800 and 1900 MHz) and T-Mobile (1900 MHz). Upgrades from GPRS to 3G will be in the direction of UMTS, often through EDGE as an intermediate step [26].

Enhanced Data rates for GSM (nee Global) Evolution (EDGE) is a 2.5G standard developed by ETSI in 1999 and touted as the final stage in the evolution of data communications within the existing GSM standards. The only IMT-2000 specification based on TDMA, EDGE supports data transmission rates up to 473.6 kbps over GSM FDD channels 200 kHz wide through an improved modulation technique. 8-Phase Shift Keying (8-PSK) involves eight levels of phase shift and, therefore, supports three bits per symbol. EDGE supports 124 FDD channels, each of which supports eight time slots/users. EDGE supports two modes of operation:

· Enhanced Circuit-Switched Data (ECSD) is an enhancement of the native GSM circuit-switched protocol. ECSD adds 8-PSK as a modulation option, thereby increasing the efficiency of data transmission. A GMSK connection requires four time slots to support a 57.6-kbps data rate, but ECSD requires only two.

EDGE also runs over IS-136 TDMA networks in the United States. In either case, EDGE is an intermediate step between 2G TDMA and 3G WCDMA, although some TDMA-based carriers may stop at EDGE. At the time of this writing (June 2007), the Cingular Wireless network (800 and 1900 MHz) largely supports EDGE and the T-Mobile network (1900 MHz) is fully upgraded [26]. Maximum transmission rates typically are in the range of 75–150 kbps.

Also known as Wideband CDMA (W-CDMA), Universal Mobile Telecommunications System (UMTS) is a 3G technology that is seen as a logical upgrade to GSM, although the two are not compatible. UMTS runs over a carrier 5 MHz wide, compared to the 200-kHz carrier used for narrowband CDMA. UMTS specifications provide for both TDD mode and FDD mode, with TDD largely used in Europe and FDD in the United States. The FDD specifications call for the downlink to run in the 2100-MHz range (2110–2200 MHz) and the uplink in the 1900-MHz range (1885–2025 MHz). As is the case with all true 3G systems, UMTS specifications include 128 kbps for high-mobility applications, 384 kbps for pedestrian speed applications, and 2 Mbps (1.920Mbps) for fixed in-building applications. In reality, UMTS currently caps the transmission rate at a theoretical 384 kbps. UMTS was first deployed in Japan (2000), where it is known as Freedom of Mobile Multimedia Access (FOMA). In some locations in the United States, Cingular has deployed UMTS running in the 850-and 1900-MHz bands. T-Mobile has committed to UMTS worldwide.

Developed by Qualcomm, the company that commercialized CDMA, Code Division Multiple Access 2000 (CDMA2000) is a 3G system based on earlier CDMA versions (also known as TIA/EIA IS-95a and IS-95b). CDMA2000 (also known as IS-856) has been approved by the ITU-R as part of the IMT-2000 family. The initial version, known as CDMA2000 1 × RTT (one times Radio Transmission Technology, with one times referring to standard channel width), offers 2.5G capabilities within a single standard 1.25-MHz channel, effectively doubling the voice capacity of the predecessor 2G cdmaOne systems and offering theoretical data speeds up to 153 kbps (throughput in the range of 70–90 kbps) through the use of QPSK modulation. An enhanced 3G version known as 1 × EV-DO (one carrier EVolution-Data Optimized) is a High-Data-Rate (HDR) version that employs 16-QPSK modulation in support of a peak data rate of 2.4 Mbps on the downlink and 153 kbps on the uplink. 1 × EV-DO supports average aggregate throughput in a fully loaded three-sectored cell of 4.1 Mbps on the downlinks and 660 kbps on the uplinks, with dynamically assigned data rates providing each user with optimum throughput at any given moment. 1 × EV-DO can run in any band (e.g., 450, 800, 1800, and 1900 MHz) and can coexist in any type of network (e.g., CDMA2000, cdmaOne, GSM, TDMA, and AMPS). CDMA2000 runs in the 800-MHz and 1.8–2.0-GHz spectrum. GSM1x is a version designed as a transition specification for GSM operators, involving dual-mode phones. Also known as IS-2000-A, 3x is an enhancement that uses three cdmaOne carriers for total bandwidth of 3.75 MHz. This supports data rates up to 2 Mbps by spreading a multicast signal over the three carriers. In North America, Bell Canada, Sprint Nextel, and Verizon Wireless (800 MHz) have committed to CDMA2000 [26, 34–40]. In August 2006, however, Sprint Nextel announced plans to invest up to US $ 3 billion on an 802.16e, aka mobile WiMAX, network that the company expects to be operational in 2008. Mobile WiMAX is based on Intel technology, rather than Qualcomm's CDMA.

While the official reason behind the TD-SCDMA is the avoidance of reliance on Western technology, some suggest that the real reason is the avoidance of patent fees and license fees to Qualcomm and others. Given the fact that there were 190 million users of GSM and CDMA cell phones in the PRC in 2002 and the then current rate of growth was 5 million per month, the PRC certainly had plenty of incentive to avoid those costs. Additionally, it apparently is quite inexpensive to upgrade a cellular network from GSM to TD-SCDMA, while it is very expensive to upgrade a GSM network to UMTS [42]. TD-SCDMA specifications have been finalized, system testing is in progress, and operational systems were expected at the end of 2006.

Contemporary cellular terminal equipment takes a number of forms today. Cellular telephones are by far the most common and are available in hundreds of styles, most of which are very small form factor models compared to the early bricks, so called because they literally were about the size and weight of a clay brick. The small form factor certainly has advantages in terms of size and weight but also presents problems with respect to the user interface, as the visual display and data entry keys are so small. These handset issues are of greatest significance with respect to high-end multifunction sets and with data, rather than voice, applications such as Web surfing and SMS messaging.

As the underlying technologies have evolved and 2.5G and 3G mobile stations have become available, cellular telephones increasingly have been oriented toward data, image, and now video applications. Voice, SMS messaging, and text-based e-mail access have been commonplace on cellular telephones since 2000 or so. Instant Messaging (IM) is much more recent. High-resolution color displays combined with high-speed network access have made Web surfing an integral part of the mobile wireless experience for several years. Digital photography and videog-raphy capabilities have now turned the high-end cell phone into an untethered digital camera. In 2004, camera phones made up 36 percent of mobile phone shipments in the United States and more than 50 percent in Europe and dominated the market in Japan [43].

By the way, cell phones do not just ring any longer, they play ring tones, and have done so for some years. Ring tones are not just the tones that you hear when your cell phone rings, they now also can be the tones that the caller hears. Just in case the several dozen ring tones that come with a typical cell phone are not to your liking, you can download additional ring tones for a charge. In fact, ring tones are now a billion dollar business in The United States.

PDAs continue to increase in popularity and ever more are wireless enabled. Similarly, laptop and tablet PCs not only can be interfaced with the cellular networks via PCMCIA cards but also now are available with built-in cellular antennas and software, right out of the box, so to speak.

While Europe and North America may be reaching saturation levels, at least for cellular voice, the developing world is largely an untapped market. The fastest growth rates are in China and Africa. It is estimated that 80 percent of the world's population lives within range of a mobile network but that only about 25 percent have a mobile phone. A good deal of emphasis these days, therefore, is on low-cost cell phones, including prepaid phones, as this combination is especially attractive to low-income users, particularly if they participate in a cash-only economy. As governments increasingly view communications as a basic human right and as a huge stimulant to entrepreneurial economies, they increasingly are lowering taxes, tariffs, and duties on them.

The trend toward cellular radio seems unstoppable, at least for voice and mobile data communications applications. It also appears that the trend will only accelerate as 3G, 3.5G, and even 4G cellular become widely available and data rates increase to truly broadband speeds. As manufacturers and carriers collaborate on multimode devices that can seamlessly bridge the gaps between Wi-Fi, WiMAX, and 3G+ cellular networks, we may very well find ourselves in a largely wireless world. By that time, of course, the definition of broadband will have ratcheted up a number of notches, and wireline networks will deliver multimedia content at Tbps speeds.

In the United States, Sprint Nextel announced in August 2006 that it would spend up to US $ 3 billion to build an 802.16e, aka mobile WiMAX, network slated to begin operation in 2008. The network will be based on technology developed by Intel, which has taken the lead in WiMAX development. Sprint Nextel, with 51.7 million cellular subscribers, expects that the mobile WiMAX network will support connection speeds about five times faster than its current PCS network, which is based on CDMA technology from Qualcomm.

The ubiquity of multifunctional cell phones and other mobile devices has clearly changed the way that many people conduct their professional and personal lives. Some years ago, you could be reasonably certain that someone walking down the street waving his arms and yelling at thin air was mentally unbalanced, but today that person is more likely just another loud, rude cell phone user with a Bluetooth-enabled headset. Many restaurants, theaters, and other venues have been declared cell-free zones in defense of these obnoxious people. (If you are one of those people and I have offended you, I sincerely hope that you will get over it—and mend your ways. )

Cellular technology is just an enabler, of course, but it has been put to some uses that are questionable, to say the least. Cell phones with built-in cameras and video recorders have been used to take upskirt or downblouse photos and for various other illicit purposes. As a result, some gyms and health clubs have banned cell phones from locker rooms. In the United States, some states have passed laws against such invasions of privacy. In December 2004, the United States enacted the video Voyeurism Prevention Act, making it illegal to sneak photos or videotapes of people for lascivious purposes on federal land, such as national parks. A study conducted by Juniper Research and quoted in The New York Times indicated that U.S. cellphone adult content was a $ 1 billion business in 2005 and predicted that it would increase to $ 2.1 billion in 2006. Camera phones have been used for good as well. The phones have been used to help catch thieves, burglars, rapists, and all sorts of criminals.

As might be expected, religious leaders have taken steps to combat, or at least control, this trend. In 2004, Saudi Arabia's highest religious authority issued an edict banning the use of camera phones, blaming them for spreading obscenity. The concern goes quite far in conservative Muslim societies, where religious authorities fear that camera phones are misused to photograph women without their knowledge and without being properly covered from view, head to toe. According to the Associated Press, a wedding party in July 2004 turned violent after a female guest was caught taking photographs with her phone—scuffles broke out and some guests were hospitalized [52]. (Note: In the previous December, the authorities banned dolls and stuffed animals, so a ban on camera phones should hardly be surprising.)

At about the same time, Mirs Communications, the Israeli subsidiary of Motorola, introduced the kosher phone, which carries the seal of approval of Israel's ultra-Orthodox rabbinical authorities. The kosher phone is stripped down to its bare functional essentials of making and receiving calls. There is no text messaging, no IM, no Internet access, no camera, and no video. Among the more than 10,000 numbers blocked are those for phone sex and dating services. Arab service providers have sought information from Mirs via backchannels using envoys sent from Jordan, which has a peace treaty with Israel [53].

Various safety issues have long swirled around cell phones and their use. Studies conducted over the years have proved and then disproved links between cell phone usage and brain cancer. The most recent of those studies, conducted at the Swedish National Institute for Working Life, indicated that those who heavily used wireless phones had a 240 percent increased risk of a cancer ous tumor on the side of the head where they used their phone. Heavy use was defined as 2000 or more hours, which roughly translates into one hour of usage a day for 10 years. Kjell Mild, who led the study, said, "The way to get the risk down is to use handsfree". Mild did not offer any suggestions with respect to the hands-free technology, although I suspect that Bluetooth might not do a lot to solve the problem. The U.S. Food and Drug Administration plans to review the study [54, 55].

Studies have long shown links between cell phone usage and vehicle accidents. It seems that driving and dialing, and even driving and talking, are distracting and that even cell phones equipped with hands-free options do not eliminate the problem. As a result, some states in the United States have passed laws banning their use while operating a vehicle. As is always the case, some people deny the link and refuse to abide by the law. (Note: They are probably the same people who deny the link between safety and seat belts. You know who you are.)

11.10 PACKET DATA RADIO NETWORKS

Cellular networks of various descriptions have supported data communications for generations, so to speak, but not very well. The 1G analog and 2G digital networks really were optimized for voice, and understandably so. CDPD offered reasonable data communications speeds, but it really was not a satisfactory solution. Alongside those cellular networks low-speed, packet-switched wireless data networks known as Mobitex (Mobile text) were deployed in Europe and the United States, and they remain in operation. Developed jointly by Ericsson and Swedish Telecom, Mobitex offers runs in Europe in the band 400–450 MHz and in the United States in the 800- and 900-MHz bands, although it will run in any band.

Mobitex channel spacing of 12.5 kHz supports theoretical data transmission rates of 8 kbps through GMSK modulation. Mobitex packets are up to 512 octets, including overhead, using a slotted Aloha protocol. Slotted Aloha is an improvement on the original Aloha protocol developed at the University of Hawaii in the early 1970s as a contention management mechanism for use in interisland wireless networks. The basic concept behind the Aloha protocol later was incorporated into the Carrier Sense Multiple Access (CSMA) protocols that Bob Metcalf and his associates developed for Ethernet.

Mobitex networks support SMS messaging, corporate e-mail access, and Internet access. Web surfing is supported through proprietary gateways that can filter out graphics or lower their resolution to reduce their bandwidth intensity. Security is provided through various proprietary authentication and encryption mechanisms. Applications include sales agents, maintenance fleets, and truck fleets. Terminal equipment supported includes two-way pagers, PDAs and various other hand helds, and laptops. Mobitex networks largely were retired in Europe in the face of strong competition from GSM but remain in place in 30 countries, including in the United States and Canada. These networks face strong competition from 2.5G and 3G cellular networks that support voice, SMS and MMS, e-mail, and Internet access from traditional cellular phones, laptop and tablet PCs, and PDAs and other hand helds.

11.11 SATELLITE SYSTEMS: LEOs, MEOs, AND GEOs

LEO and MEO systems are configured as constellations of small, low-power satellites. In combination, the satellites in such a constellation generally provide full coverage of major land masses, and some have been designed to provide full coverage of virtually every square inch of the earth's surface. The various proposals have included as many as 840 satellites and are intended to provide various combinations of voice and data services. These systems also are known as Mobile Satellite Systems (MSSs), as opposed to the Fixed Satellite Systems (FSSs) in geostatic orbit. Whizzing around the earth like electrons whizzing around the nucleus of an atom and as illustrated in Figure 11.11, LEO and MEO networks are designed so that a satellite is always within reach of a terrestrial terminal.

Propagation delay is not much of an issue for several reasons. First, the satellites are in low-earth orbits of approximately 485 miles, so the uplink and downlink propagation delays are not significant. Second, the satellites communicate directly over switched intersatellite links, so only one uplink and one downlink are required. As the satellites whiz around the earth like electrons whiz around the nucleus of an atom, however, they have a very short dwell time, so you cannot maintain contact with any given satellite for very long, and neither can the other party involved in this telephone call. Therefore, the satellite with which you established the connection must pass that call off to another satellite before it gets out of view, and so must the satellite at the terminating end, and so must every satellite in between. Think of it as a switched cellular network in reverse. In a cellular network, the antennas at the cell sites are stationary. As you whiz through the cellular network in your high-speed vehicle, the cellular network maintains the connection through a hand-off process between cell sites. In an Iridium environment, you are the one who is (relatively) stationary, while the cell sites whiz around you. Iridium also, by the way, supports pagers and airplane communications via the same L-band frequencies. If you call a device that is not on the Iridium network, your call can be connected to the existing PSTN and cellular networks via 12 regional gateways.

Now, I may not be as smart as the Teledesic folks, and I certainly do not have pockets nearly as deep as the least rich of the lot, but I bet my next royalty check from this book that the Teledesic portal will not be optimized for Netscape Navigator. As I said before, it is an incredible idea. Note, however, that Teledesic was conceived in 1986, when there were no optical fiber backbones, no DSL, no cable modems, no PON, and no WiMAX and when Internet access typically involved modems that ran at the blazing speed of 19.2 kbps, best case. In October 2002, Tele-desic suspended construction work, and in July 2003, the company gave up its frequency licenses. So, another great idea conceived by engineers went kerplooey.

In the meantime, Craig McCaw and a subset of the principals of the failed Tele-desic venture had moved on to ICO, for Intermediate Circular Orbit, which they rescued from bankruptcy in 2000. The company reemerged as New ICO, and the system was redesigned. New ICO is a MEO design slated for 10 operational satellites in two orbital planes. After a failed launch or two, New ICO faded into the background for several years. As recently as April 2005, however, New ICO announced that Space Systems/Loral was awarded a contract to build a geostationary mobile satellite system. So, it appears that the system was redesigned yet again and is now a GEO rather a MEO design.

Inmarsat, for International Maritime Sat ellite, is the pioneer in the MSS space, having been in service since the early 1990s. Inmarsat also is based on a bent-pipe GEO design, with no intersatellite links. Each Inmarsat GEO generates 19 wide spot beams and 228 narrow spot beams. The system reportedly covers 98 percent of the world's population with services including voice, e-mail, Internet access, fax, and videoconferencing at speeds up to 492 kbps [66].

11.12 AND THAT'S NOT ALL

The creativity and inventiveness of the modern human mind should not be underestimated, for there is still more technology to explore, although some of it may seem a bit odd, if not downright wacky. Remember, however, that the telephone seemed like a wacky idea to many people just 130 years ago. Please read on.

Along similar lines, Sanswire Networks plans to offer a variety of wireless services via high-altitude airships known as Stratellites. The airships will take up positions in the stratosphere at 65,000 ft (approximately 13 miles, or 20 km) and hold to a single GPS coordinate through the use of solar-powered electrical engines. Each Stratellite will support a payload of several thousand pounds and have clear Line-Of-Sight (LOS) coverage of about 300,000 square miles, an area roughly the size of Texas, according to company information [70].

And there is more. Angel Technologies Corporation plans to develop a high-capacity wireless metropolitan network dubbed HALO (High-Altitude Long Operation). The network will use shifts of specially designed aircraft capable of carrying a telecommunications payload of 2300 lb and remaining aloft for 8h or so at an altitude of approximately 51,000 ft. At that altitude, the coverage area is about 75 miles in diameter. HALO is expected to provide raw aggregate capacity exceeding 16 Gbps, in support of 50,000–100,000 customers requiring symmetric T1 packet data rates [71].

Chapter 12: Video and Multimedia Networking

"Ouch! You struck ‘me,’ you mean!" replied Ned, rubbing his shoulder, where the young inventor had imparted a resounding blow of joy.

"What of it?" exclaimed Tom. "My apparatus works! I can send a picture by telephone! It's great, Ned!"

"But I don't exactly understand how it happened," said Ned, in some bewilderment, as he gazed at the selenium plate.

"Neither do I," admitted Tom.

Victor Appleton, Tom Swift and His Photo Telephone or The Picture That Saved a Fortune, Gross & Dunlap, 1914

On the whole, you may wonder how critical interactive video and multimedia networking are to our daily lives. The answer is not clear. While we all understand that a picture can be worth a thousand words and we clearly benefit from one-way video and multimedia, the ability to engage in an interactive video or multimedia communications is not necessarily that important to most of us, either personally or professionally. Such interactive networking currently is expensive in terms of both bandwidth and equipment, although those costs certainly will come down over time. Many of the highly touted benefits have yet to be proven, although a compelling case is developing.

Certainly, video and multimedia have application in education. People tend to have different learning styles. Some are visual learners and need to see something in order to process the information. Some are auditory and need to hear about it. Still others tend to be more kinesthetic and need to touch it. In any event, the addition of the visual information stream certainly assists in the educational process, as it does in the communications process, in general. A fascinating application is in the world of health care. As discussed later in this chapter, telemedicine has proved itself to be of great importance in the delivery of high-quality medical care to remote areas where specialists, and even general practitioners, may not be available.

12.1 VIDEO COMMUNICATIONS: DEFINED AND EVOLVED

The first public demonstration by wire and wireless of Television or "Distant Seeing," as developed by the staff of the Bell System, took place on April 7, 1927. Participating in the demonstration at Washington, D.C., and New York were notable gatherings of leaders in the fields of science, industry and public affairs. Those who talked from the Bell Laboratories in New York were able to see plainly the features of those in Washington with whom they conversed over the long distance circuits of the Bell System. [This] … was followed by demonstration of Television by radio in which the audience in New York saw the artists visualized on the screen and heard a varied program from the radio experimental station of the Laboratories at Whippany, N.J.

Things Worth Knowing about the Telephone, American Telephone & Telegraph Company, 1931

The development of modern TV largely was due to the efforts of Herbert E. Ives, a scientist at Bell Telephone Laboratories. As the son of Frederick Eugene Ives, who in 1878 developed the first practical process for making halftone printing plates, Ives was oriented toward the visual world. In 1923, he and his associates combined the photoelectric cell with the vacuum tube repeater to produce the first commercial system for the rapid transmission of pictures over telephone wires for application by the daily press. Using a scanning beam developed by Frank Gray, another Bell Labs scientist, multiple real-time images of people were transmitted in rapid succession, and TV was born. The first public demonstration, on a black-and-white TV, was conducted in April 1927; color TV first was demonstrated in June 1929 [3]. That experimental color transmission included the transmission of pictures of an American flag, a watermelon, and a bunch of roses. The transmitting system had three sets of photoelectric cells, amplifiers, and glow-tubes, with each filtering out one color—Red, Green, or Blue (RGB). At the receiving end, mirrors superimposed the monochromatic images to create a single color image [1]. In the 1930s, the first commercial TV stations began operation over the radio waves.

Coaxial cable entered the world of TV in 1936, when the first experimental transmission took place between New York and Philadelphia. Jointly conducted by AT&T and the Philadelphia Electric Storage Battery Company (Philco), the experi-ment proved highly successful in terms of transmission performance, as multiple frequencies could be transmitted over the same shielded medium. In 1950, AT&T opened the first coaxial cable for coast-to-coast TV transmission [4]. Community Antenna TeleVision (CATV), generally based on coaxial cable, largely has sup-planted broadcast TV in the United States. The popularity of cable is due in large part to the large number and wide variety of available channels; broadcast TV simply cannot compete at this level due to spectrum limitations. More recently, CATV spread to a number of other developed countries, including England and Australia.

Now, only two key evolutionary concepts remain—those of the transmitters and receivers. Vladimir Zworykin, a Russian immigrant to the United States, built on his graduate work in Russia where he had studied the nature of fluorescence under Boris Rosing. On January 1, 1939, Zworykin received patents for his iconoscope (transmitting) and kinescope (receiving) tubes, which formed the basis for modern cameras as well as Cathode Ray Tubes (CRTs)used as display devices in traditional TV sets and computer monitors [1, 5].

12.2 VIDEO BASICS

In order to fully comprehend the nature and implications of advanced video communications, it is necessary to understand the basics of video. The basic concepts are frame rate, scanning, resolution, aspect ratio, luminance, chrominance, and synchronization.

· Frame rate refers to the rate at which frames of still images are transmitted. Video is a series of still images that are transmitted in succession to create the perception of fluidity of motion. If transmitted in rapid succession, the perception is one of complete fluidity; 24 frames per second (fps) is considered to be motion picture quality, and 30 fps is considered to be broadcast quality. If the frames are transmitted at a slow rate, the result is a poor-quality, herky-jerky video that creates a strobe-light effect reminiscent of 1970s discos. Particularly below 15 fps, quality suffers quite noticeably, as the fluidity of motion is lost even though the image quality may be quite high.

· Scanning refers to the process of refreshing the screen; the scanning rate is a function of the power source of the receiver. Interlaced scanning, which is used with most analog TV systems, involves two fields. Odd lines (field 1) are refreshed in one scan and even lines (field 2) in the next. Each set of odd and even lines refreshed constitutes a frame refreshed. For example, the American NTSC standard provides for 30 fps, involving 60 scans, which relates directly to the 60 Hz of the U.S. power source. The European PAL standard provides for 25 fps, involving 50 scans, which relates directly to the 50 Hz of the European power source. Progressive scanning involves displaying all horizontal scan lines in one frame at the same time, which avoids the problem of interline flicker. While the refresh rate varies, most contemporary PC monitors operate at 60-90 fps. The difference between the two techniques is imperceptible to most of us, until we see a broadcast TV news show that includes a video clip of PC monitors.

· Resolution refers to the definition, or sharpness, of the image. Resolution is determined by the number and areal density of the pixels, or pels (picture elements), which essentially are dots of picture similar to the dots in half-tone printing. The greater the number and density of the pixels, the better the resolution. If the same number of pixels is spread over a greater area, the result is a grainier picture, as you can readily see by sitting close to a big-screen TV.

· Aspect ratio refers to the relationship between the width and the height of the image. The 4: 3 (4 wide to 3 high) aspect ratio specified by the American NTSC standard is rooted in the early days of TV, when round picture tubes made effective use of this approach.

· Chrominance refers to color, with different standards permitting varying levels of color depth. Clearly, the video image is more pleasing and lifelike when the range of color is as broad as possible.

· Synchronization between the transmitter and receiver includes vertical and horizontal synch, both of which are critical. Vertical synch is required to keep the picture from scrolling, or flipping. Horizontal synch keeps the picture from twisting.

12.3 ANALOG TV STANDARDS

Television standards are several and incompatible. The initial standards were set in the United States, where broadcast TV originated. In 1945, the Federal Communications Commission (FCC) set the initial VHF (Very High Frequency) transmission standards at 4.5 MHz. The National Television Standards Committee (NTSC) was formed in 1948 to standardize the characteristics of the broadcast signal. Ultimately, the Radio Corporation of America (RCA), which was owned by AT&T, lobbied the Electronics Institute of America (EIA) and set the initial black-and-white TV standards. Color TV was commercialized some years later. Among the first live color TV broadcasts were the Cotton Bowl and Rose Bowl football games on January 1, 1954 [1]. As the cost of color TV sets was quite high, color television did not really become popular until the 1960s when sets became affordable for the masses. Table 12.1 compares the major analog standards of NTSC and PAL, which are discussed in more detail in the bulleted list that follows, with digital HDTV.

· National Television Standards Committee (NTSC) was established in the United States as the first standard (1953), setting the tone for broadcast TV. NTSC is defined in ITU-R Recommendation 1125. While other standards have since been developed, they all derived from the NTSC baseline. NTSC is characterized as analog in nature, with 525 interlaced scanlines. There are 640 pixels per line, 485 of which are dedicated to the active picture. The frame rate is 30 fps, 60 fields interlaced, and the aspect ratio is 4: 3. An early analog standard that is viewed as overly complex and ineffective in the contemporary digital world, NTSC also is said to mean Never The Same Color [6].

· Séquential Couleur Avec Mémoire (SECAM) (which translates as sequential color with memory), a variation of PAL, was developed in France. In addition to its use in France, it also is the standard in regions once under French influ-ence, including areas of the Middle East.

12.4 DIGITAL TV AND HIGH-DEFINITION TV

Digital TV (DTV) transmission offers the same advantages as any other form of digital communications, including enhanced bandwidth efficiency through compression, improved signal quality, and more effective management and control. Additionally, digital video content must be recognized as nothing more than video data. As a result, it offers advantages in terms of processing, storage, and manipulation. Those advantages include editing, alteration (e.g., morphing), reproduction, compression, and store-and-forward capability.

TV and most video, for that matter, in their native forms comprise synchronized analog image and analog voice information. Digital video requires that the information be digitized through the use of a video codec. A broadcast-quality video signal is extremely bandwidth and storage intensive. For example, the resolution of a digital video might require 640 horizontal and 480 vertical pixels. As chrominance and luminance require 24 bits per pixel, the video signal, alone, requires 7,372,800 bits per frame. At a frame rate of 30 fps, the bandwidth requirement is 221,184,000 bps. Clearly, it is not within the realm of reason to store and transmit a broadcast-quality video signal requiring 221+ Mbps. It is clear, therefore, that compression is critical if the networks are to support digital TV.

Despite all the advantages of digital technology as applied to TV production, storage, transmission, and reception, all of the TV sets were analog. A conversion from analog to digital broadcast TV is revolutionary and completely unthinkable, unless there exists a compelling reason for the changeout of untold millions of TV sets. According to the manufacturers and the FCC, that compelling reason is in the form of High-Definition TV (HDTV).

Considering that an HDTV video signal at 1080 × 1920 pixels, 24 bits per pixel, and 30 frames per second requires bandwidth of almost 1.5 Gbps, compression becomes very important. The ATSC standard specifies the compression algorithm as MPEG-2, which is discussed later in this chapter and the transport subsystem as ISO/IEC 13818. Packet transport involves a serial data stream of packets of 188 octets, 1 octet of which is a synchronization byte and 187 octets of which are payload. This packet approach is suitable for Asynchronous Transfer Mode (ATM) switching, as each 188-octet MPEG-2 packet fits nicely into the payload of four ATM cells, with only 4 octets of padding required. Forward Error Correction (FEC) is employed in the form of Reed–Solomon coding. Radio Frequency (RF) modulation is accomplished using 8 Vestigial Sideband (8 VSB), which supports a bit rate of 19.28 Mbps over a 6-MHz terrestrial broadcast channel through the use of eight discrete amplitude levels. Audio compression is based on the AC-3 specification from Dolby Digital and the ATSC. The audio sampling rate is 48 kHz, and the system supports six channels in the Dolby Digital surround format. That format specifies multiple channel outputs, including center, left and right center, left and right surround, and Low-Frequency Enhancement (LFE), also known as subwoofer [7]. Despite the fact that most programming is not yet in HDTV format and that most broadcast and cable signals are still analog, Forrester Research estimates that about 15.7 million households had HDTV sets by the end of 2005 and predicts that number to be 50 million by 2009 [9].

Each station also is responsible for the costs associated with upgrades to digital transmission facilities. The costs of converting the several hundred million analog TV receivers are the responsibility of the viewers, although beginning in 2008, the National Telecommunications and Information Administration (NTIA) will provide to eligible households up to two $ 40 vouchers good toward the purchase of digital-to-analog converters to retrofit existing analog TV sets, Video Cassette Recorders (VCRs), and other devices. New TV sets, with screens from 25 to 36 in., were required to have built-in digital tuners as of March 1, 2006. By March 2007, all new TVs, regardless of size, and other devices designed to receive broadcast TV signals must have digital tuners built in. According to U.S. census projections, there were approximately 110 million households in the United States at the end of 2005 [10, 11]. Considering the fact that the average household had 2.6 television sets, the costs of adding 286 million analog TV sets and 100 million or so VCRs to existing landfills will be substantial. Assuming the use of a single landfill, just imagine its size. Just imagine the bewilderment of future archeologists when they excavate it. (This assumes that there will be future generations, which is questionable, since the mercury and lead that leaches out of the TV sets and out of the landfill into the streams, rivers, lakes, and oceans of the world may have killed off all life as we know it.) They will probably think it was some sort of sacrifice to the TV gods, and they will not be far off. In any event and as a byproduct of the conversion from analog to digital broadcast TV, the 750 MHz of analog spectrum will be reclaimed for fixed and mobile commercial wireless services, and that spectrum will be auctioned once the reclamation is complete [12, 13].

You may wonder why I spent so much time discussing broadcast TV technology and standards and their evolution. The reasons are several. First, most of the same technologies and standards apply to interactive videoconferencing and multimedia. Second, although broadcast DTV and CATV do not necessarily have to follow the same rules of the road, it is likely that the CATV providers will carry broadcast TV signals in a digital format since they are upgrading their networks to digital anyway. The FCC has imposed must-carry rules on both the CATV and the DBS (Direct Broadcast Satellite) providers, thereby forcing them to carry local broadcast TV channels and to fairly compensate the local stations for that content. Considering that service providers are deploying broadband digital local loops in support of the triple play of voice, data, and TV, that the backbones are digital, that data, image, and movie content are digital, and that the TV broadcast signal will be digital, it is pretty clear that the digital circle will be complete very soon.

12.5 BANDWIDTH AND COMPRESSION

Digital video, on the other hand, can be compressed fairly easily. Because uncompressed, broadcast-quality digital video requires between 90 and 270 Mbps, compression is critical. Without compression, a 1-Gbps fiber-optic network could accommodate no more than 11 digitized NTSC channels. To digitize and compress the video information stream, the analog video and data signal first must be digitized through the use of a codec. Clearly, it is possible to reduce the amount of bandwidth required to transmit digital video and the amount of memory required to store it by reducing the frame rate, resolution, or color depth. However, the result is less than pleasing. In order to maintain the quality of the video presentation, therefore, the data must be compressed by using an appropriate and powerful data compression algorithm.

Lossless compression enables faithful reproduction of the video signal, with no data loss, although compression rates are in the range of only 10: 1. Lossy compression tends to produce artifacts, which are unintended and unwanted visual aberrations in the video image that often show up as jagged blockings or tiling effect known as aliasing, banding of colors, white spots, and even dropped frames. Although these artifacts result in a degraded picture, this lossy approach achieves compression rates up to 200: 1 [14]. Actually, compressed video currently can be transmitted with quite acceptable quality at T1 speeds of 1.544 Mbps or less. MPEG, for example, uses lossy compression in the form of Discrete Cosine Transform (DCT). A number of steps are involved in video compression, including filtering, color-space conversion, scaling, transforms, quantization and compaction, and interframe compression.

· Filtering, also known as image decimation, reduces the total frequency of the analog signal through a process of averaging the values of neighboring pixels or lines. For example, adjoining black-and-white pixels become gray pixels. Taps are the number of lines or pixels considered in this process; MPEG, for example, uses a seven-tap filter.

· Color-space conversion, also known as color sampling, simply involves the reduction of color information in the image. As the human eye is not highly sensitive to slight color variations, the impact is not noticeable. Black and white, however, are prioritized because the human eye is very sensitive at that level.

· Scaling addresses the creation of the digital image according to the presentation resolution scale associated with the display device. Rather than digitizing the video signal in large scale, the codec is tuned to the scale of presentation in terms of horizontal and vertical pixels, thereby reducing the amount of data that must be digitized. In consideration of this factor, the aspect ratio must be standardized.

· Quantization and compaction encoding simply reduce the number of bits required to represent a color pixel. Compaction techniques include run-length encoding, Huffman coding, and arithmetic coding.

· Interframe compression considers and eliminates redundant information in successive video frames. The background of a movie scene, for example, might not change, even though the actors move around the set. While the motion of the actors must be reflected, the background need not be retransmitted over the network. Rather, it can be compressed out of each frame until background changes must be reflected. At the point of decompression at the receive end of the signal, the unchanged background can simply be reinserted over and over again.

12.6 VIDEO STANDARDS

According to an anonymous but very wise man, "The nice thing about standards is that there are so many from which to choose!" Video compression standards are no exception—they are numerous and incompatible. Early standards were developed for specific purposes, such as motion picture production (MPEG) and photographic editing (JPEG). The ITU-T became involved in the development of international standards toward Broadband-ISDN (B-ISDN). In the videoconferencing world, numerous proprietary standards have been developed that require that the network involve equipment from only a single vendor.

As of late, a number of formal standards have developed. Those video compression standards of significance include P × 64, JPEG, and MPEG and its variations. Virtually all manufacturers of significance have embraced these standards, incorporating one or more of them into their systems alongside a proprietary compression technique. Generally involving P × 64, this approach provides at least a minimum level of communication between systems of disparate origin. The world of desktop videoconferencing is much more parochial; proprietary solutions long predominated, yielding incompatibility all too often [15]. As always is the case in networking, the highest common denominator rules. Lacking a common denominator, anarchy rules. The common denominator currently is P × 64, a standards-based solution that virtually all manufacturers support.

Actually, H.261 is an element of the ITU-T H.320 umbrella standard for video telephony over circuit-switched ISDN and addressing narrowband visual telecommunications systems and terminal equipment. Related ITU-T standards include H.221, which defines a frame structure in support of audiovisual teleservices in 64-kbps channels, and H.222, which defines the frame structure for such services in an ISDN environment. While these umbrella standards truly are international in nature, the U.S. and European camps unfortunately (and not surprisingly) are divided over certain implementation aspects of the ITU-T standards. Specifically, those differences deal with the manner in which audio and still-frame graphics are handled. For graphics, for example, the Europeans have adopted JPEG (ITU-T T.81), which is compatible with most PC-based graphics software. That approach is totally incompatible with the H.261 standard implemented in the United States, although it does provide a bridge to true multimedia applications [16]. H.320 is discussed in greater detail later in this chapter.

The ISO and ITU-T jointly developed Joint Photographic Experts Group (JPEG) as a compression standard for editing still images as well as color facsimile, desktop publishing, graphic arts, and medical imaging. A symmetrical compression technique, JPEG is equally expensive, processor intensive, and time consuming in terms of both compression and decompression. Motion JPEG is used in the editing of digital video. JPEG is not appropriate for video transmission, as the compression ratio is in the range of only 20: 1-30: 1. Therefore, JPEG transmission in support of videoconferencing requires bandwidth in the range of 10-240 Mbps, which is far too bandwidth intensive.

MPEG standards are several and still in final development stages. MPEG standards provide very high compression levels and excellent presentation quality. MPEG is a joint technical committee of the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC). MPEG offers the critical advantage of asymmetric compression and decompression. While MPEG compression is time consuming and expensive, the decompression process is rapid and involves relatively inexpensive equipment. MPEG compression is as high as 200: 1 for low-motion video of VHS quality, and broadcast quality can be achieved at 6 Mbps. Audio is supported at rates from 32-384 kbps for up to two stereo channels.

· MPEG-2, defined in November 1994 as ISO/IEC IS 13818, is the standard for DTV. MPEG-2 was conceived as a broadcast standard for interlaced images at 720 × 480 pixels at 30 fps and requiring bandwidth at 4-100 Mbps over transmission facilities (e.g., fiber optics, hybrid fiber/coax, and satellite) capable of such support. While MPEG-2 requires much more bandwidth than MPEG-1, it provides much better resolution and image quality and at much greater speed. MPEG-2 has found application in Direct Broadcast Satellite (DBS) services, also known as Direct Satellite Systems (DSS). Such services employ Ku-band satellites and Very Small Aperture Terminal (VSAT) dishes in competition with CATV, running MPEG at rates of about 3 and 7.5 Mbps [16]. In a convergence scenario, MPEG-2 is the standard of choice, supporting compression rates as high as 200: 1, depending on content specifics. MPEG-3, designed for HDTV application, was folded into MPEG-2 in 1992.

· MPEG-4 Version 1 was approved in October 1998 as ISO/IEC IS 14496 and as the standard for multimedia applications. Version 2 was approved in 1999. MPEG-4 is a low-bit-rate version intended for application in broadcast TV, videophones, and mobile phones and other small hand-held devices. In a wireless context, MPEG-4 is designed for IMT 2000 applications that transmit at a rate of up to 384 kbps upstream and receive at a rate of up to 2 Mbps downstream, although rates of 64 kbps and lower are supported. Encompassing both client/server and mass storage-based playback scenarios, MPEG-4 deals with the coded representation of audiovisual objects, both natural and synthetic (e.g., computer generated), and their multiplexing and demultiplexing for transmission, playback, and storage.

· MPEG-7, approved in September 2001, officially is known as the Multimedia Content Description Interface. MPEG-7 is intended to be the content representation standard for multimedia information search, filtering, management, processing, and retrieval. MPEG-7 essentially is metadata standard based on XML (eXtensible Markup Language) for describing multimedia content features in order that one can easily search for multimedia content on the Web. (Note: According to the ISO, there apparently is no known reason for the failure to expand the logical numbering sequence and, instead, to jump to MPEG-7 and then to MPEG-21.)

12.7 INTERNET PROTOCOL TELEVISION (IPTV)

Several forces have concerned to create a great deal of interest in TV distribution over IP networks. There is an undeniable and irreversible movement towards IP, in general. Telephone companies are aggressively deploying broadband local loop technologies in the local loop to support triple-play services comprising voice, Internet access, and TV. While copper-based ADSL has limited ability to support TV, VDSL certainly can support it over short distances. The telco trend is more towards optical fiber in the loop, either directly from the Central Office (CO) to the premises in a Fiber-To-The-Premises (FTTP) Passive Optical Network (PON) scenario or a hybrid Fiber-To-The-Neighborhood (FTTN) scenario that terminates fiber in a neighborhood node and uses ADSL or VDSL over a short length of embedded copper cable from the node to the premises. Even full-rate ADSL supports enough bandwidth for at least one compressed video channel. VDSL and PON both support multiple channels.

The telco IPTV approach differs considerably from that of traditional CATV networks. CATV networks support analog TV over 6-MHz channels allocated from the total 850-MHz spectrum typically supported over coax or hybrid fiber/coax cable plant. The number of channels clearly is limited to the number of 6-MHz RF channels that can be derived from the available spectrum. As CATV networks are configured as point-to-multipoint networks, each channel is delivered to every connected subscriber premises. The IPTV approach delivers each channel only to the subscriber premises that select to view it, whether one or hundreds do so. Over a PON local loop, the TV channel occupies a separate wavelength from the head end to the premises. In an FTTN scenario, the IPTV signal is converted from optical to electrical format at the neighborhood node and travels over Unshielded Twisted-Pair (UTP) copper cable plant the last few hundred meters or so. AT&T's Project Lightspeed, for example, calls for 20–25 Mbps to each of 300–500 subscriber premises connected to its fiber-optic loops. Verizon's FiOS PON network offers 5, 15, and 30 Mbps, depending on subscriber preferences. Either network offers bandwidth sufficient to support multiple DTV and even HDTV channels, properly compressed. At the moment (9:05 A.M. PDT, May 21, 2007), however, the telco networks run video in uncompressed analog format over a separate wavelength at 1550 nm. The eventual conversion to IPTV speaks to compressed DTV, in packet format, likely still but not necessarily over a separate wavelength.

12.8 THE H.320 FAMILY OF MULTIMEDIA STANDARDS

· Gateways are optional elements in the H.323 environment used for various levels of protocol conversion. A gateway comprises a Media Gateway (MG) and a Media Gateway Controller (MGC), which commonly are physically distinct devices. The MG serves as a protocol converter between devices and networks that have native H.323 capability and those that do not. The gateway also may translate between audio, video, and data formats. Finally, the gateway may perform signaling conversions between the H.225 packet protocol and external protocols such as Signaling System 7 (SS7) and Q.931. Alternatively, signaling conversions may be performed by gatekeepers, call processors, or session border controllers.

· Multipoint Control Units (MCUs) support conferencing among three or more participating terminals. The MCU comprises a Multipoint Controller (MC) and optional Multipoint Processors (MPs). The MC is responsible for call control negotiation to achieve common levels of communication. The MP may process either a single media stream or multiple media streams, depending on the nature of the conference.

H.324 is the ITU-T Recommendation for low-bit-rate multimedia communication over the analog PSTN through V.34 modems. As such, modems are limited to maximum transmission rates of 28.8 kbps. Voice must be highly compressed in order to make room for video and other visual information streams [17–20].

The ITU-T Recommendation for the multipoint transport of multimedia data is T.120, which data can include whiteboarding or binary files. This series of recommendations supports a broad range of underlying network technologies and can work either alone or under the H.320 umbrella. T.120 is entirely platform independent and can run in a variety of network environments, involving either reliable or unreliable data transport. Unicast and multicast modes both are supported [17,21].

12.9 SESSION INITIATION PROTOCOL

The Internet Engineering Task Force (IETF) defined Session Initiation Protocol (SIP) in its RFC 2543 (March 1999) as an application layer (Layer 7 of the OSI Reference Model) signaling protocol for establishing, modifying, and terminating multimedia sessions or calls over an IP network. Finally approved in June 2002 in RFC 3261, which obsoleted RFC 2543, SIP is a modular component of IP telephony, although it can function over any network. SIP offers considerable advantages over H.323, although it can be used in conjunction with H.323 where appropriate. H.323 commonly is criticized as being too slow in establishing sessions, that is, setting up packet calls. An H.323-compliant client initiating a call must query a gatekeeper for the address of a new destination device. Once the gatekeeper has provided that address, the originating client establishes the session by using the H.225 signaling protocol, and the two clients negotiate features and call control procedures by using H.245. All of this takes time. In fact, it can take much more time than is required to set up a call through the conventional circuit-switched PSTN. The exact amount of time is sensitive to network specifics, and you must consider the fact that the Internet is a network of many networks, each of which is of uncertain nature in many respects. H.323 also is criticized for being overly complex and highly centralized.

Things Worth Knowing about The Telephone, American Telephone and Telegraph Company, 1931

Videoconferencing systems consist of cameras, monitors, video boards, microphones, speakers, and software. Videoconferencing can be accomplished in the workplace over a LAN. While it generally is more effective, easier, and less costly to walk across the hall and hold a face-to-face meeting, contemporary LANs often extend across multiple floors and even multiple buildings in a campus setting. Videoconferencing over the WAN clearly offers great benefits in terms of reduced cost and increased availability for meetings. Systems for videoconferencing can be quite substantial or can be PC based. Videoconferencing has increased significantly over the past few years as the cost of equipment and bandwidth have decreased—systems are made up of room systems, rollabout systems, and PC-based systems.

· Room systems are complex and quite expensive systems intended for videoconferencing among groups of people. While a specially designed room and equipment easily could cost $ 250,000 in the 1980s, a room system can be configured for less than $ 25,000 today. AT&T, MCI, and Sprint have provided room-based conferencing services for years from select locations, and they also provide network-based videoconferencing services for large corporations. Such service offerings include access and transport services as well as network-based MCUs. While many U.S. Incumbent Local Exchange Carriers (ILECs) in the late 1980s or early 1990s announced plans for video dial tone, those announcements clearly were way ahead of both the technologies and the market, and all of those plans were abandoned. The proliferation of fiber optics in the network core and advent of PON in the local loop now make such a service technically feasible, at least in terms of transmission capacity. Kinko's Copy Centers (now FedEx Kinko's) began deploying videoconferencing services in 1994 and currently offers such services at 150 locations in the United States and Canada. Kinko's has improved quality to 30 fps at 384 kbps through an alliance with Sprint for high-speed access and transport.

· Rollabout systems essentially are portable and much less expensive versions of room systems and also are intended largely for group-to-group conferencing. Such systems account for the preponderance of the stand-alone equipment market.

12.12 VIDEOCONFERENCING EQUIPMENT

Videophones originated with the AT&T Picturephone, which was demonstrated at the New York World's Fair in 1964. Never intended for practical application, the Picturephone was extremely bandwidth intensive, requiring bandwidth of about 90 MHz and weighing about 26 lb [6]. During the 1980s, AT&T, BT, and others developed videophones that sold for less than $ 1000. As the cost was high, as each party was required to have a videophone of the same manufacture, and as the picture quality was poor (2 fps), videophones were stunning failures.

Regardless of the nature of the system, some combination of hardware and software is required. Videoconferencing equipment includes transmit (camera) and receive (display) equipment that operate in concert with and through various intermediate devices to format the signal properly and otherwise treat it for effective transmission over a network. Those intermediate devices include codecs, inverse multiplexers, servers, and control units. Figure 12.3 presents a simple videoconferencing arrangement.

· Inverse Multiplexers (inverse MUXs) are used in commercial videoconferencing systems where dedicated bandwidth is not available for relatively bandwidth intensive communications. An inverse MUX splits the video signal into two or more component parts that are transmitted over separate circuits or, perhaps, separate channels of multiple multichannel circuits (e.g., T1). The inverse MUX on the receiving end reassembles and resynchronizes the complete video signal for proper presentation.

· Servers are extremely high capacity storage devices, containing many gigabytes or even terabytes of memory. Servers store video and audio data for delivery to clients on demand.

· Multipoint control units are digital switching and bridging devices that support multipoint videoconferencing, with up to 28 parties (sites) commonly supported. MCUs must be compatible with the compression standards employed with the codecs. H.231, for example, describes ITU-T MCU standards, and T.120 describes generic data conference control functions. MCUs may be found in the carrier network in support of a carrier videoconferencing service or on the end-user premises in support of a videoconferencing network based on leased lines.

12.13 WAN VIDEOCONFERENCING NETWORKS

Video networking can be accomplished over a number of facilities and service offerings, depending on the application and the amount of bandwidth required. Analog circuits will support videoconferencing at low speeds, although the results are less than completely pleasing. The failed videophones offered by AT&T, BT, and others made use of dial-up analog circuits or ISDN circuits. As always is the case, digital circuits offer better performance than do analog circuits.

ISDN circuits are preferable to analog circuits because they provide more bandwidth and better transmission quality. However, the higher cost and lower availability of ISDN have slowed the acceptance of videophones based on ISDN technology in the United States. Note that in Europe and Japan ISDN is not particularly expensive and is much more widely available. Switched 56/64-kbps circuits can be used for videoconferencing—generally aggregated or bonded to provide multiple channels. Switched 384-kbps connectivity can be provided on the basis of fractional DS-1 through ISDN PRI channels in a channel group known as HO or over ADSL.

DS-1 facilities support full-motion, high-quality videoconferencing over dedicated networks at rates up to 2.048 Mbps for E-1 and 1.544 Mbps for T1. However, such facilities are costly and not widely dedicated to such applications. Large user organizations with dedicated leased-line T/E-carrier backbone networks make highly effective use of videoconferencing; the video communications contend with voice and data for network access through intelligent MUXs.

12.14 VIDEO OVER IP

I spent a good deal of ink in several previous chapters examining VoIP. Now, I want to shift the emphasis to the other VoIP, Video over I P, which is quite different from IPTV. Video, like voice, is stream oriented and isochronous in nature. That is to say that video, like voice, depends on a continuous flow of interrelated data from the transmitter across the network to the receiver. For that matter, video, in general, usually is a combination of video and voice, and I cannot think of an example of videoconferencing that is not characterized by the same combination. The video and voice information streams are layered together and carefully synchronized so that the voice audio matches the movement of the speaker's lips. Just like stand-alone voice, interactive videoconferencing is very demanding in terms of QoS. It is critical that latency is in the range of 150 ms or less, that jitter be in the range of 50 ms or less, and that packet loss be in the range of 1 percent or less. Issues of latency and jitter are not particularly critical in a one-way streaming video application, as the receiver has time to buffer the incoming data stream and adjust. In a real-time, two-way videoconference, however, these issues are very significant. That's quite a trick over an IP network, but it can be done—and increasingly is being done with quite satisfactory results.

That said, it may surprise you—and it certainly surprised me—to hear that one of the reasons behind the increase in IP videoconferencing is that it actually works better than ISDN. According to Chris DiFiglia, president of the Polycom User Group, the typical 384-kbps ISDN videoconference depends on getting six bonded ISDN B channels to work simultaneously, which can be difficult. DiFiglia states that "IP is a much more fault-tolerant network. With ISDN, if there's a dropped frame or picture the entire call is dropped. An IP world is more forgiving. The worst that would happen is that the video frame would freeze for a second" [31]. IP videoconferencing certainly is less expensive, as ISDN is notoriously costly, and the six B channels carry per-minute charges, which can be particularly significant for international calls. ISDN videoconferencing also requires fairly expensive software and terminal equipment or terminal adapters. I P, on the other hand, generally involves little additional cost.

12.15 MULTIMEDIA CONFERENCING

Software must be in place to support electronic text, image, audio, and video information in a multimedia conference. While the voice and video aspects of the conference are supported in a fairly straightforward manner, the real and distinct advantage of multimedia conferencing is that of enabling multiple parties to collaborate on textual and graphic documents. Special software enables each party to contribute to such documents, in collaboration with the other parties—hence the term collaborative computing. During such a collaborative session, the original text document is saved, while each party contributes changes that are identifiable as such, by contributor. Once the parties agree to the collaborative edits and enhancements, the entire text file is refreshed and saved.

Similarly, a design or a concept can be developed graphically and on a collaborative basis through whiteboarding, much as the parties would do on a physical whiteboard in a face-to-face meeting. Typically, each party to the conference has access to a special whiteboard pad and stylus which is used to draw. Each party can modify the initial drawing, with each individual's contribution identified by separate color. Again, and once the group has agreed on the final graphic rendition, the graphic is saved and all screens are refreshed.

There can be little doubt about the value of enhancing a communication with visual information. Pictures add another and very important element to the process of learning and comprehension. Moving pictures enable us to see the person to whom we are talking, thereby creating a more natural and effective person-to-person communication. Taking it one step further, a truly collaborative effort is enhanced greatly when multiple persons can work on a document. Supporting all of this over a network can save a lot of money, time, shoe leather, and gasoline. In a February 2004 report, Yankee Group analysts cited a number of benefits for videoconferencing, particularly in a contemporary multinational, multicultural environment [32]:

· Fifty-five percent of the impact of communications comes from facial expressions and body language, and only 38 percent from vocal inflection. (Source: UCLA.)

· Attendees in face-to-face meetings retain 38 percent more information than those in audio-only meetings. (Source: Harvard University and Columbia University.)

· Face-to-face meetings increase the power or persuasion by 43 percent over audio-only meetings. (Source: 3M Company.)

· Attendees learn 200 percent more in face-to-face meetings than in audio-only meetings. (Source: University of Wisconsin.)

· Attendees absorb information as much as 40 percent faster in face-to-face meetings than with audio alone. (Source: Wharton School of Business.)

Add to these benefits of electronic face-to-face meetings the hard dollar savings in travel expense and employee travel time, and a compelling case can be made for electronic conferencing. The unfortunate events of September 11, 2001, added the element of personal safety to the list of forces driving renewed interest in videoconferencing. Serving to supplement, but never replace, face-to-face in-person collaboration, video and multimedia networking has a legitimate place in the networked world.

The development of videoconferencing has been much slower than many of us projected. Traditional system and network technologies never developed to the point that high-quality videoconferencing was either simple or affordable. The recent development of standards for IP video promises to overcome those issues of complexity and cost, given the high availability and reasonable cost of broadband capacity in both the core and the local loop. Large organizations just cannot deny its value, and the costs of the system technologies are becoming quite reasonable for video-intensive applications. As costs continue to drop, small-business, home business, and even consumer markets for multimedia equipment will expand greatly.

Another interesting application for videoconferencing is in the justice system. Judges around the country, including in my little town of Mt. Vernon, Washington, make use of videoconferencing systems for video arraignments and other court appearances required for suspected criminal defendants. The clear advantage is that a video communication between the judge and lawyers in the courtroom and the defendant in the jailhouse eliminates the costly and sometimes dangerous process of transporting the accused and convicted. Tom Swift would have been proud, indeed, of the evolution and application of his photo phone.

Yet another interesting potential application is that of voice-and video-enabled websites. Clearly, the customer's contact experience can be improved tremendously if he or she can click a videoconference button and connect to an agent to establish a videoconference. That assumes, of course, that both parties are properly prepared for the camera.

Despite increased bandwidth, improved compression, and all the various protocols and mechanisms that have improved, and will continue to improve, video quality over the years, the videoconferencing experiencing still leaves us flat—two dimensional, that is. Wouldn't if be nice if video were in a 3-D (three-dimensional) format? Well, there has been a lot of development effort expended to do just that, and there is at least one commercially available system. Teleportec has developed a system that creates a 3-D effect through the use of multiple cameras on the transmit side and on the receive side by projecting the image on a specially treated sheet of glass that acts as a beam splitter, refracting light at different rates. The system builds on the H.320 and H.323 standards and currently runs over ISDN or dedicated circuits. A full T1 is preferred for maximum quality, but 768 and 384 kbps provide acceptable quality [33].