Trends in the communications industry spring from advances in semiconductors.
The telecommunications network has been converting its technology from conventional analog transmission to a digitized network in order to improve voice communication. The industry is now preparing for the influx of data communications it will be expected to handle in the near future.
The digital network climinates most of the interference characteristics that plague long-distance analog transmission, greatly improving the quality of service. Other advantages for this conversion are mainly economical. Serviceability, power consumption and minimized read estate all help reduce the cost of operating a network and therefore encourage rapid digital conversion.
The digital central-office shipments forecast for 1984 were very accurate, as approximately 14 million lines were installed. The
Bell operating companies installed 9,349 switches in 1984; 433 of these were digital. For 1985, the projections showed an overall reduction of switch installations to 9,075. However, the digital switch installations were projected to increase to 1,029--almost three times that installed in the previous year. Data communications has the potential to have enormous worldwide socio-economic impact, surfacing primarily as increased productivity through automation of engineering, manufacturing and office activities.
As more people join the data processing revolution, the need to transmit and receive data over the public network increases. The traditional data communications product for public network use has been the modem--the device that converts digital data into an analog format before transmission. This technique is still the most-adopted method of transmission and is likely to stay that way for some years to come, especially for the single terminal user. Small business and private subscriber networks will use modems throughout this decade and well into the next, transmitting data either via an internal PBX or the public network at data rates between 300 and 9600 baud, although 1200 and 2400 baud will be the most popular:
The cost of this form of transmission will continue to drop as further integration of semiconductors takes place. Even as the modem we know gives way to ther forms of data networking, modem techniques will be used in other more-exotic forms. But before this happens, the integrated modem will become a very sophisticated functional block almost as popular as the telephone set is today.
Modems will have facilities for data compression that will increase data-handling capabilities. Forward error control and correction will reduce bit-error-rate conditions. They will change transmission speed automatically, depending on line conditions. For security, they will contain encryption techniques. Interfacing with the public network, the modem will contain internal "feature phone" functions such as auto-dial, auto-answer, redial, number directory, automatic interface to a host computer and a digital message system.
To perform all these functions, the modem would need to operate at five million instructions a second with two million multiples, equivalent to about 30 IBM PCs. The modem can be used in many applications but is generally too slow for sophisticated computer-to-computer communications or complex data processing needs.
Aside from the modem, there are many different approaches to high-speed data transfer. These cover a myriad of local-area network (LAN) solutions to techniques that integrate both voice and data under the banner of the integrated services digital network (ISDN). There are also the ultimate-speed requirements using optical fibers, with addressing speeds in the 100 to 200 megabit-per-second range.
The telecommunications and data processing industries are convinced that the Information Age will merge the two functions, producing the largest business segment in the industrial world. The race is on. Major players from both industries have strategic plans for a piece of the information and automation pie. The large central-office switch manufacturers are for the first time as technologically advanced as the PBX manufacturers, and are aggressively integrating voice and data into their switches. The thrust behind ISDN is single station-to-station voice/data transfer. It allows approximately six times the data rate of a high-speed modem--more than adequate for buman interface. It also makes use of the existing telephone cabling schemes, reducing the investment cost to industry. The private branch exchange is changing significantly in order to support the voice/data requirements. The next PBX generation will include integral LANs that are intrinsically part of the PBX architecture, providing the potential for fully integrated wideband data transmission as well as voice and messaging communications. The PBX will have the capability of allocating bandwidth dynamically, depending on the nature of each message. Packet channels will be integral to the system, allowing easy access of data terminals to the switch. A layered software structure, combining switch software and computer software within the system, will provide for integration of message and active call processing. Also, these new systems will be fully distributed. Each node can provide fully transparent communications with any other node in the system, no matter where it is.
The New PBX's Advantages over the LAN
This new PBX system will have advantages over the more-traditional LAN, which consists of a single cable to which all devices are attached. A particular device wishing to communicate must gain access to this medium while all others will have to wait for the medium to be relinquished. The PBX is inherently capable of establishing and maintaining multiple simultaneous voice and data calls. Consequently, the cumulative data rate and bandwidth of a voice/data PBX is much greater than a LAN. The PBX will also act as a gateway between other LANs and WANs operating in the same environment. The PBX will therefore become the hub of a factory and office automation program directing human and computer communications both internally through many layers and externally via the public network. Most of these facilities will be made possible through continuous use of semiconductors, increasing levels of integration far beyond what we see today.
The growth for these new switches looks promising. It will not, however, be fueled by economics alone, but will be facility-driven and therefore customer-dependent. Technology will not be the gating item--customer needs will. This usually means slower but more-consistent growth as the new integrated capabilities are accepted by business and industry.
The main thrusts in the data processing and communications industry are factory and office automation programs. Automated processes produce large amounts of data that have to be transported throughout the system for analysis and response. A lot of this data is being produced in real time, which requires high-speed data networks connecting large decision-making administrative computer systems.
Computers absorb and manipulate data much faster than humans. Data communication should not be the weak link. Ideally, transportation of data should be faster than the central processor can deal with it. Different networks have specific advantages and disadvantages when applied to specific situations, indicating that many different networks will co-exist within a given environment. These data-only networks come in many forms, usually under the banner of local-area networks or wide-area networks, often called "value-added." LANs are usually baseband systems, in which the transmitting signal on the cable is the data itself reprsented by line changes. WANs are broadband systems where the transmitting signal employs multiple high-frequency carrier waves dispersed across the transmission medium's bandwidth. The carrier can be modulated by frequency, phase or amplitude.
Topology and Medium Dependent on Volume, Distance
The three most-commonly adopted topologies are bus, ring and star. The transmission medium can be twisted-pair telephone wires, coaxial cable, fiber optics, radio or microwave. The topology and medium are dependent on data volume and distance requirements. Many different systems will co-exist, taking many hardware and software forms. A total network system could be compared to a road and highway system crossing the country. At the lowest level are driveways and other roadways that lead to local streets. These join larger tributaries. The country roads lead to faster-paced highways. These intersect with even-larger freeways and interstate highways that crisscross the country. So it is with data communications--the highway systems contain stop and yield signs, traffic lights, entrance and exit ramps, acceleration lanes and cloverleafs that permit smooth transition from one road to another, or one network to another.
Linking networks of dissimilar format requires a high-speed protocol converter (gateway). This can be a mini-computer or a PBX using complex computer architecture. The use of new VLSI circuits with integration levels far greater than that of today will allow a cost-effective method of performing these functions. The growth for this industry will be dependent on the adoption of factory and office automation programs. The projections show good growth through the end of this decade and will increase substantially during the next.
The factory/office automation customer requires that all data and voice communications be controlled from one source and not the variety of telephones and data terminals that now exist. The term being used for this terminal is the integrated voice/data terminal (IVDT). The IVDT will include voice communications, "feature phone" functions and data terminal capability. Its software would be based on an office desktop or factory floor. A feature phone, personal computer and workstation will be combined into this terminal. It will also act as a gateway between local LANs and the internal communications network.
These units will become as prolific as the telephone, first in the office and then at home. But before the IVDT is a fully integrated system, there will be many intermediate products taking form somewhere between the standard telephone and a full-function workstation. The next five years will see dramatic growth for feature phones. The features will be very simple to very complex, integrating complex computer voice/data message functions. As with the PC market, the initial thrust will be into an office environment, and from there to the vast residential market.
Even though the EIA is predicting an overall reduction in phone sales for 1984 to 1985, the feature phone market is expected to expand dramatically in 1986. In 1984, feature phone only represented three percent of the total phone market, allowing room for substantial growth. The system supporting feature phones and IVDTs will also mature, producing vast amounts of data through a company's network.
Many companies have multiple locations requiring that this data be transmitted over large geographic areas, either via a private network or the public network. These networks must handle many times the data volume of that at any individual site. The three most-popular forms of high-speed transmission are packet-switching Tl links (today's public channels) passing data at 1.5 to 2.0 Mb/s, fiber-optic cables at 50 to 200 Mb/s or microwave and RF cellular systems. As an example it would take 10 Tl links to connect two IBM mainframe computers and transfer data in real time.
Fiber Optics to Be Developed as a Long-Term Strategy
It seems unrealistic that Tl links will be adopted as a long-term strategy for large data users. Fiber optics are seen as the revolutionary concept for transmission, linking major cities and local centers using packet switching and high-speed data techniques. The number of switching levels in the system can be reduced significantly, creating an economic reason for the change.
Fiber optics use two forms of light transmission--laser and gallium-arsenide diodes. While lasers will be used for long-haul transmission, gallium arsenide, at a significantly reduced cost, will be used in many applications for factory and office commonications alike.
Apart from the high data rates possible, fiber optics have excellent environmental qualities. Fiber cable transmits information as light and neither generates nor is affected by electromagnetic interference (EMI). This makes it ideal for use in high-EMI environments such as factory floors. Lightwaves experience relatively low signal loss, resulting in efficient communications over tens of kilometers without repeaters. Standard protocol specifications are being produced by the American Nation Standards Institute, and will be known as the fiber distributed-data interface (FDDI) X3T9.5. This interface uses token passing and time/token transmission protocols and a 4B/5B encode/decode technique to reduce system bandwidth as a relationship to data rates. Although these are new techniques, it is expected that large-volume production will be in place by the end of this decade with strong growth during the next.
The business world is a mobile environment. Personal communications will have to be more portable. The concept being adopted is cellular radio. The cellular concept is ideal for mobile communications such as two-way radio, paging systems and, of course, mobile phones. By the end of this decade, it is expected that as with fiber optics, there will be cellular highways interconnecting ground and air locations. There will be cellular satellites connecting cell ground stations, which in turn will connect to the public telecommunications network, establishing total communications coverage. The cellular industry is still in its infancy, but as it matures, mobile communications products will increase in complexity. Cellular-radio digital paging using cellular modems will act as a portable message center. Two-way radio dispatch can be mobile to mobile, mobile to base, and mobile to telephone or IVDT. Automobile phones will be voice/data units, voice activated for hands-on-the-wheel safety.
It is conceivable that by the 21st century, pocket communicators will be commonplace. These voice/data units would report on home environmental systems and message communications between home and business. Being a two-way system, remote control of these functions can be activated via the mobile unit. These advances will be driven by market requirements and not technology, although technology will continue to reduce cost, creating a large user base. If the concept is adopted, there eventually could be as many mobile units in operation as there are residentail telephones in use today.
Telecom Growth and New Kinds of Semiconductors
Man's success has been affected most by his ability to communicate. Technological advancements have been escalating exponentially for some time. The communications industry is one that will change dramatically over the next decade or two, changing the lifestyle of almost everyone, everywhere. Most of this capability will be brought about through VLST circuitry producing true "systems on a chip." The telecom and data processing industries are by far the largest users of semiconductors.
"Merchant" or independent semiconductor manufacturers are generally affected more strongly by tends in the telecom industry than by any of their other markets.
Thus, the semiconductor industry is right now responding aggressively with new, more-complex devices. It too is seeing the need to automate processes using sophisticated, computer-controlled functions networked together. Producing more-efficient, cost-effective manufacturing is one of the reasons the whole data processing communication program started. The market for equipment that automates semiconductor production is expected to grow rapidly in the coming years, climbing from $334 million in 1984 to over $1 billion in 1988.
The communications industry invested $1.3 billion for test equipment in 1983, and is expected to increase that investment at a compound annual growth rate of 18 percent, reaching $3 billion by 1988.
The major suppliers of this equipment are Hewlett-Packard and Tektronix, presumably through the sale of benchtop instrumentation. It can be seen that ATE manufacturers such as GenRad and Teradyne are beginning to get a piece of the action. As the complexity of functions within the system increase, testing strategies must also change. The switch systems are becoming fully integrated data processing and transport systems using more and more VLSI technology. This can be seen by the increasing use of integrated circuits within the systems. The communications system designers use linear, digital and mixed signal devices, being by far the largest user of linear/digital parts. The semiconductor and board test industries are also affected by the need for more-sophisticated testing. The new mixed-technology semiconductors now pose the same testing problems at the board, assembly and system level.
The Joining of Computer and Communications Markets
In the US, major players AT&T, IBM and Northern Telecom are all actively converting system technology IBM is, without question, the world leader in data processing, while AT&T is the largest manufacturer of switching, microwave, fiber-optic transmission, channel banks and modem equipment. Both companies have produced most of their own semiconductors. This is now changing, as the variety of semiconductor types used increases. Moving toward market integration, IBM and AT&T have acquired interests in each other's fields. IBM has acquired expertise in the communications industry through partnerships with Rolm (PBX), SBS (satellite communications), Motorola (cellular), CBS (videotec), Merrill Lynch (financial-information systems) and MCI (long distance). IBM is projecting a market penetration of between $10 billion to $15 billion in telecommunications by 1988. In 1984, it only had a five percent share of $45 billion. It has a long way to go. Many of its products are noncompatible and have many diferent operating systems. It still has to establish its networking strategy, and it has very little communication experience. AT&T, on the other hand, has established joint ventures in the computer industry with Olivetti (PC), Convergent Technologies (PC), Amdahl (software) and an agreement with General Motors to supply software for its factory automation programs.
AT&T has more of an integrated system than IBM. Its Unix operating system will most likely become the standard over the next 10 years. It has networks and experience the industry needs. The main weakness is its marketing ability in the data processing industry, a lack of data products and installed base.
As the market becomes larger, many new companies seek entrance by concentrating on system rather than semiconductor design, although custom semiconductor processes are becoming very popular--especially in the data processing industry, where proprietary design can delay duplication.
The US, by virtue of its technological advancements, is the largest market for communications and data processing equipment. Although Europe and Japan are both technologically advanced, the market growth in these areas has been restricted by government control. As this control is relinquished, both geographic regions will expand exponentially, penetrating new geographic and market areas.
The Far East is, at this time, dominated by Japan. Its internal and export plans for communication advancements are aggressive. Internally, Nippon Telephone and Telegraph (NTT), as a private company, is aggressively converting Japan's communications network from the more-traditional to its equivalent of ISDN--the information network system (INS). The INS almost exclusively parallels the CCITT's ISDN formats for voice packt data and facsimile. The INS has provisions for digitization, single network fiber and satellite transmission, video, machine-to-machine interface and enhanced communications processing. The goal is to have a nationwide fiber-optic cable and satellite INS by 1995. During 1985, NTT will release a contract for $16.6 billion to digitize all circuits.
The Far East is potentially the largest overall growth market for new subscriber lines, rather than replacement lines as in Europe and the US. Far East nations will be spending large amounts on the need to communicate: China, $3.7 billion over the nest six years; India, $12.5 billion over the next five years; Thailand, one million lines forecasted to be installed.
South Korea's aggressive communication plans include manufacturing and export. The two largest companies in the Southern Hemisphere for communications equipment are NEC and up-and-coming Gold Star. NEC's advantage is that approximately 30 percent of its business is communication--30 percent data processing and 30 percent semiconductors--the perfect combination for factory office automation expertise.
At $8 billion, NEC certainly has the resources to be a formidable player in the IBM--AT&T arena. NEC exported $450 million in products into the US in 1983, $800 million in 1984 and an estimated $1 billion in 1985. NEC is expected to continue to grow at around 25 percent per annum.
International Boundaries Are More Like Walls
Europe, on the other hand, does not have such a clear direction. Its international boundaries act more like walls. The communications industry must establish consistent strategies between countries if the major European manufacturers stand a chance of competing against US and Far East companies. Conversion from analog to digital switching systems is the most advanced in Europe at 29.7 million lines in service by the end of 1984. Some 33 percent of these were Alecatel-Thompson, Northern Telecom, 20 percent; and L.M. Ericsson, 15 percent.
Europe, with all its international problems, will still spend around $8 billion a year for the next 10 years on public switching systems. In 1984, France spent $4 billion, 15-percent increase on 1983. Germany spent $5.3 billion, Britain, $3.0 billion, and Italy, $2.5 billion. The telecom industry in Europe represents the largest market for semiconductor manufacturers, expected to grow from more than $1 billion in 1984 to more than $7 billion by 1993--$6 billion being integrated circuits.
In summary, it can be seen that the advances in the data processing/telecom industry over the next two decades will be significant. Geographically, large markets exist at all four corners of the globe. Communications has no boundaries.
|Printer friendly Cite/link Email Feedback|
|Date:||Mar 1, 1986|
|Previous Article:||Developer-Reseller sees growth potential as customers join private switched network.|
|Next Article:||Communications, interconnectivity drive office automation strategies.|
|Industrial production and capacity utilization: annual revision and 1997 developments.|
|ATHEROS LAUNCHES SINGLE-CHIP 802.11G FOR PCI EXPRESS.|