In the previous section we traced the evolution of communication networks from telegraphy to the emerging integrated services networks. Before proceeding with the technical details of networking, however, we pause to discuss factors that influence the evolution of communication networks. Figure 1.19 shows the three traditional factors: technology, regulation, and market. To these we add standards, a set of technical specifications followed by manufacturers or service providers, as a fourth factor.
A traditional axiom of telecommunications was that a new telecommunications service could succeed only if three conditions were satisfied. First of all the technology must be available to implement the service in a cost-effective manner. Second, government regulations must permit such a service to be offered. Third, the market for the service must exist. These three conditions were applicable in the monopoly environment where a single provider made all the decisions regarding design and implementation of the network. The move away from single providers of network services and manufacturers of equipment made compliance with recognized standards essential.
The availability of the technology to implement a service in and of itself does not guarantee its success. Numerous failures in new service offerings can be traced back to the nontechnology factors. Frequently new services fall in gray areas where the regulatory constraints are not clear. For example, most regula-
tory policies regarding television broadcasting are intended for radio broadcast and cable systems; however, it is not clear that these regulations apply to television over the Internet. Also, it is seldom clear ahead of time that a market exists for a given new service. For example, the deployment of videotelephony has met with failure several times in the past few decades due to lack of market.
Technology always plays a role in determining what can be built. The capabilities of various technologies have improved dramatically over the past two centuries. These improvements in capabilities have been accompanied by reductions in cost. As a result, many systems that were simply impossible two decades ago have become not only feasible but also cost-effective.
Of course, fundamental physical considerations place limits on what technology can ultimately achieve. For example, no signal can propagate faster than the speed of light, and hence there is a minimum delay or latency in the transfer of a message between two points a certain distance apart. However, while bounded by physical laws, substantial opportunities for further improvement in enabling technologies remain.
The capabilities of a given technology can be traced over a period of time and found to form an S-shaped curve, as shown in Figure 1.20a. During the initial phase the capabilities of the technology improve dramatically, but eventually the capabilities saturate as they approach fundamental limitations. An example of this situation is the capability of copper wires to carry information measured in bits per second. As the capabilities of a given technology approach saturation, innovations that provide the same capabilities but within a new technology class arise. For example, as copper wire transmission approached its fundamental limitation, the class of coaxial cable transmission emerged, which in turn was replaced by the class of optical fiber transmission. The optical
FIGURE 1.20 Capability of a technology as an S curve (based on Martin 1977)
Second type of invention
Third type of invention
FIGURE 1.20 Capability of a technology as an S curve (based on Martin 1977)
fiber class has much higher fundamental limits in terms of achievable transmission rates and its S curve is now only in the early phase. When the S curves for different classes of technologies are superimposed, they form a smooth S curve themselves, as shown in Figure 1.20b.
In discussing the evolution of network architecture in section 1.2, we referred to the capability curve for information transmission shown in Figure 1.9. The figure traces the evolution from telegraphy to analog telephony, computer networks, digital telephony, and the currently emerging integrated services networks. In the figure we also note various milestones in the evolution of networking concepts. Early telegraphy systems operated at a speed equivalent to tens of bits per second. Early digital telephone systems handled 24 voice channels per wire, equivalent to about 1,500,000 bits per second. In 1997 optical transmissions systems could handle about 500,000 simultaneous voice channels, equivalent to about 1010 bits per second (10 gigabits per second)! In the year 2000 systems can operate at rates of 1012 bits per second (1 terabit per second) and higher! These dramatic improvements in transmission capability have driven the evolution of networks from telegraphy messaging to voice telephony and currently to image and video communications.
In addition to information transmission capacity, a number of other key technologies have participated in the development of communication networks. These include signal processing technology and digital computer technology. In particular, computer memory capacity and computer processing capacity play a key role in the operation of network switches and the implementation of network protocols. These two technologies have thus greatly influenced the development of networks. For more than three decades now, computer technology has improved at a rate that every 18 to 24 months the same dollar buys twice the performance in computer processing, and computer storage. These improvements have resulted in networks that not only can handle greater volumes of information and greater data rates but also can carry out more sophisticated processing and hence support a wider range of services.
It should be noted that the advances in capabilities in technology do not just happen. They are due to the creativity of numerous engineers and scientists that work relentlessly in pursuing these advances. It should also be noted that advances in the core technologies are not the only factor in providing advances in technology. The development of new algorithms to design, control, and manage large systems are key to handling the complexity associated with modern systems. Indeed, most of this book is dedicated to presenting the key concepts that are used in the design, control, and management of modern communication networks.
Traditional communication services in the form of telephony and telegraphy have been government regulated. Because of the high cost in deploying the requisite infrastructure and the importance of controlling communications, gov ernments often chose to operate communications networks as monopolies. The planning of communication networks was done over time horizons spanning several decades. This planning accounted for providing a very small set of well-defined communication services, for example, telegraph and "plain-old telephone service" (POTS). These organizations were consequently not very well prepared to introduce new services at a fast rate.
The last three decades have seen a marked move away from monopoly environment for communications. The Carterfone decision by the U.S. Federal Communications Commission (FCC) in 1968 opened the door for the connection of non-telephone-company telephone equipment to the telephone network. The breakup in 1984 of the AT&T system into an independent long-distance carrier and a number of independent regional telephone operating companies opened the way for further competition. Initially, customers had a choice of long-distance carriers. More recently, with the development of wireless radio technologies and advances in cable TV systems, competition is now possible in the access portion of the network, which connects the customer to the main telephone backbone network. Indeed the Telecommunications Act of 1996 opens the way for a far wider participation of established and new companies in the development and deployment of new services and new technologies.
The trend toward deregulation of telecommunications services has also taken place on an international basis. Countries such as the United Kingdom, New Zealand, and Australia in particular have experimented with novel approaches to providing telecommunications services in a competitive environment. As indicated above, the development of new technologies, wireless radio in particular, has enabled businesses to provide complete telecommunications services using entirely new technology infrastructures. In the third world and in countries with emerging industries, these wireless technologies have a cost advantage over the traditional wire-based systems and hence facilitate the introduction of competition by enabling the deployment of new alternative service providers.
In spite of the trend towards deregulation, telecommunications will probably never be entirely free of government regulation. For example, telephone service is now considered an essential "lifeline" service in many countries, and regulation plays a role in ensuring that access to a minimal level of service is available to everybody. Regulation can also play a role in addressing the issue of which information should be available to people over a communications network. For example, many people agree that some measures should be available to prevent children from accessing pornography over the Internet. However, there is less agreement on the degree to which information should be kept private when transmitted over a network. Should encryption be so secure that no one, not even the government in matters of national security, can decipher transmitted information? These questions are not easily answered. The point here is that regulation on these matters will provide a framework that determines what types of services and networks can be implemented.
The existence of a market for a new service is the third factor involved in determining the success of a new service. This success is ultimately determined by a customer's willingness to pay, which, of course, depends on the cost, usefulness, and appeal of the service. For a network-based service, the usefulness of the service frequently depends on there being a critical mass of subscribers. For example, telephone or e-mail service is of limited use if the number of reachable destinations is small. In addition, the cost of a service generally decreases with the size of the subscriber base due to economies of scale, for example, the cost of terminal devices and their components. The challenge then is how to manage the deployment of a service to first address a critical mass and then to grow to large scale.
As examples, we will cite one instance where the deployment to large scale failed and another where it succeeded. In the early 1970s a great amount of investment was made in the United States in developing the Picturephone service, which would provide audio-visual communications. The market for such a service did not materialize. Subsequent attempts have also failed, and only recently are we starting to see the availability of such a service piggybacking on the wide availability of personal computers.
As a second example, we consider the deployment of cellular radio telephony. The service, first introduced in the late 1970s, was initially deployed as a high-end service that would appeal to a relatively narrow segment of people who had to communicate while on the move. This deployment successfully established the initial market. The utility of being able to communicate while on the move had such broad appeal that the service mushroomed over a very short period of time. The explosive growth in the number of cellular telephone subscribers prompted the deployment of new wireless technologies.
Standards are basically agreements, with industrywide, national, and possibly international scope, that allow equipment manufactured by different vendors to be interoperable. Standards focus on interfaces that specify how equipment is physically interconnected and what procedures are used to operate across different equipment. Standards applying to data communications between computers specify the hardware and software procedures through which computers can correctly and reliably "talk to one another." Standards are extremely important in communications where the value of a network is to a large extent determined by the size of the community that can be reached. In addition, the investment required in telecommunications networks is very high, and so network operators are particularly interested in having the choice of buying equipment from multiple, competing suppliers, rather than being committed to buying equipment from a single supplier.
Standards can arise in a number of ways. In the strict sense, de jure standards result from a consultative process that occurs on a national and possibly international basis. For example, many communication standards, especially for telephony, are developed by the International Telecommunications Union (ITU), which is an organization that operates under the auspices of the United Nations. Almost every country has its own corresponding organization that is charged with the task of setting national communication standards. In addition, some standards are set by nongovernmental organizations. The TCP/IP protocol suite and the Ethernet local area network are two examples of this type of standard.3 De facto standards arise when a certain product, or class of products, becomes dominant in a market. For example, personal computers based on Intel microprocessors and the Microsoft Windows operating system are a standard in this sense.
The existence of standards enables smaller companies to enter large markets such as communication networks. These companies can focus on the development of limited but key products that are guaranteed to operate within the overall network. This environment results in an increased rate of innovation and evolution of both the technology and the standards.
On a more fundamental level, standards provide a framework that can guide the decentralized activities of the various commercial, industrial, and governmental organizations involved in the development and evolution of networks.
Was this article helpful?