Thothink

For interactive TV (ITV) networks to be universally available as the Public' s next mass medium, they will require the most economical telecomputing method from the perspectives of consumers, operators, and programmers. Only a telecomputer system based on telephone keypad signalling can meet this criteria for the successful nationwide deployment and market acceptance of interactive television networks as the Public's access to the National Information Infrastructure (NII).

EXECUTIVE SUMMARY

The telecommunications and computing industries have captured the imaginations of the Public with visions of an interactive multimedia network future. The leaders in these converging industries are each championing an interactive network model that is in the perceived best interest of their particular industry. However, sooner or later, economic realities, basic engineering, marketing, and regulatory imperatives will dictate a telecomputing architecture and approach for the "Public Network" that is quite different from today's highly publicized and so-called interactive TV trials. The following are those imperatives and realities. Only a headend or central office-based telecomputer system that is independent of the set-top box can offer... - ever more aesthetically appealing graphics and system intelligence, - transparent and unobtrusive system maintenance and upgrades, - an open systems approach; the economics of a system that can... - be traffic engineered based on customer demand, - serve all wired and wireless modes of video delivery; the creative and artistic freedoms of a system... - not restricted by hardware limitations. All such centralized telecomputer systems require a network of I/O channels for consumers to send input signals and receive audiovisual output. Only a telecomputer system based on the existing public switched telephone networks (PSTN) and NTSC broadcast video standards will offer... - the most economic I/O channels to and from consumers that can be traffic engineered - - the efficiency of I/O channels consistent with the asymmetrical nature of interactive programming - artistic freedoms not restricted by bandwidth limitations, memory constraints, nor audience reach - competition between wired and wireless modes of interactive TV program delivery - a way to trial broadband interactive video programs without having to upgrade the distribution plant for digital video or two-way traffic Although opinions may differ on ITV network architectures, one point seems to be generally accepted. That is, consumer friendly ITV networks will use simple and intuitive graphical user interfaces (GUI). When ITV programs originate from centrally located computers and broadcast their GUIs as ordinary TV video channels, we need only to provide an adequate supply of video channels and a means for consumers to remotely signal those computers. A touch-tone phone, as an ITV network's remote signally unit, is all the Public will need to start interacting. Consumers just call over the PSTN to an ITV program that converts their DTMF telephone keypad signals to a directional and triggering signalling protocol that multimedia computers can understand as point & click input. Eventually, as cable systems approach video-on-demand (VOD) capacity, such computer input devices as; joysticks, mice, and keyboards will also generate DTMF signalling and be equipped with RJ-11 jacks for the consumers' convenience. It is important to recognize that VOD means, in ITV terms, Computers-On-Demand! The TV screen will become the consumers' PC monitor. The bottom line is that ordinary touchtone phones provide the most immediate and economical means to bring interactivity to the medium of broadcast television in a way that will guarantee universal access to our interactive TV future.

INTRODUCTION

Before 1993, interactive TV was not a very popular industry topic and anyone championing such an idea, yet alone a particular network architecture or vision for delivering it, was likely met with skepticism or worse. In spite of this, a few of us still felt that an interactive medium, something more than movies-on-demand, was going to fill in the successive blank in the electronic media progression; telegraph, telephone, radio, television, _____________. Whatever this new medium was going to be, one thing was certain. The telephone, broadcast, computer, and publishing industries were going to be up to their commercial and technological necks in it. Now its 1995 and the environment is quite different. Not only has anybody that's anybody in wide variety of industries jumped on the interactive multimedia superhighway bandwagon, but they have also been struck by their own personal visions of this future. Understandably, these visions are conveniently or subconsciously biased in favor of the technologies, culture, and corporate interests of the originating visionary. Unfortunately, none of these visions take an engineering and commercial approach that is in the best interest of the Public at large. This paper approaches interactive television and telecomputing opportunities from a perspective that views all our existing and converging telephone, cable, broadcast, and computer networks as one Public Network. We need to take the macro engineering and artistic perspectives on this new medium that are free of the biases imposed by the current industrial paradigm if we are to have a next generation Public Network as a worthy successor to the Bell System.

WHAT IS A TELECOMPUTER SYSTEM?

A telecomputer system, in the fuzzy macro context of the information superhighway or the NII, is that mass media system that our nation's cable and telephone companies are striving to create to deliver interactive digital everything to the consuming Public. It is that underlying hardware infrastructure that will integrate telecommunications, television, computers, and electronic publishing into a seamless national multimedia network. There are two major telecomputer models competing to be this next generation's predominate Public Network system. The older and more familiar one is the PC in every home connected to an on-line network or "PC" model. Although this model was a dismal failure as a mass medium during the brief videotex era, it has had a recent resurgence encouraged by flashy multimedia PCs, GUIs, and the explosive worldwide growth of the Internet. The champions of this model tend to be the manufacturers and sophisticated users of advanced PCs, workstations, and high-speed data networking gear. The essential NII elements that this model brings to a telecomputer architecture are those that offer the most artistic, creative, and communication freedoms to users, programmers, and publishers. The other, getting off to a very dubious start, is the smart TV set-top box (TSB) interfacing with a video server or "TSB" telecomputing model. This model is just emerging due to recent advances in microprocessor, video compression, and network transmission technologies. The champions of this model tend to be the manufactures of cable converters, microprocessors, and mid-range computers in partnership with our cable and telephone companies. In apparent conflict with the PC model, the essential NII elements of the TSB network architecture are those necessary for responsible mass media broadcasting and network operator control. More important for ITV and the NII, the TSB telecomputer model recognizes the utility potential of an unlimited number, preferably in common carrier form, of video channels to the consumer. These two models expose most of the essential NII elements perceived as necessary for a successful new mass medium. Between the two, it is possible to extract the inherent objective of a telecomputer system. That being, to offer all consumers the potential and opportunity to access all multimedia publications over a Public Network in a responsible and socially acceptable manner. Although each of the incumbent models has technical and philosophical advantages over the other, neither will pass the test of being economically feasible as a mass medium for reasons discussed later. As a result, this paper is introducing and championing a new telecomputer model that merges the PC and TSB network models in a way that will permit the emergence of a new mass medium providing universal access to the NII.

CREATE THE IDEAL PC MODEL AND EXPLODE IT

From the perspective of a new Public Network infrastructure, the challenge facing our industry is to create a telecomputer system that can stand the test of time measured in generations of Americans. Let's start by creating the ideal PC model and identify it's economic implications for a mass media solution. The core elements of the PC telecomputer model are the user's input and output interfaces, the microprocessor and it's memory, and interfaces to the external world. To understand the paradigm shift that will occur in order for mass media convergence to take place, contemplate these core elements individually and their economic impact on the engineering of a nationwide multimedia network. For the user's input interface, the PC model provides the user with a means to send control and input signals to the computer. The user generates these signals via several familiar computer input devices. In the mass media environment, however, it is generally accepted that consumers will be using a hand-held point & click device with a minimum of push-buttons for interacting with intuitive GUIs. It is also generally envisioned over the longer term that as voice recognition technology advances, these signals will likely include voice input and commands as a replacement for the qwerty keyboard. As for the economic impact on ITV; input devices are inexpensive, require little or no maintenance, and are rarely rendered obsolete. It is on the output side of the PC model where we are now seeing and hearing the most change. Virtually the whole essence of multimedia deals with the presentation of an advanced audiovisual output interface to the consumer. The technical advances and dropping costs in computing power has led to the creation of graphics images that are indistinguishable from photographic ones. As recent high-tech movies have vividly shown, computer programming is beginning to look and sound good enough to watch in the theatre and on TV. It is this potential for attracting an audience that has Hollywood excited. The added potential for user interaction and audience participation is what has Siliwood excited. However, for all the excitement and hype, the given mass media output interface to the consumer is still only their existing TV screens and speakers. This remains true even though advances in consumer electronics may one day lead to an HDTV or other advanced NTSC standard home theatre system. Regardless, for an ITV approach based on NTSC broadcast video standards, the output side of the PC telecomputing model adds no costs over passive TV. At the center of the PC model is the microprocessor and it's memory. In the ideal PC model, the consumer will continually need the latest and fastest microprocessors with an expanding memory. Closely coupled with the microprocessor are the interfaces to the external world. First, is the access to external data storage media such as diskettes and CD-ROMs. The ideal PC model will also continually need the latest and fastest network interfaces to access on-line multimedia publications and connect with global network sources. Of course, in the ideal PC telecomputer model, there are no incompatibility nor obsolescence issues. Unfortunately, the potential for such an ideal PC model ever existing in the real world are nil. As a result, for a mass medium, the economic impact associated with these portions of the PC telecomputer model are so enormous they constitute it's fatal flaw. But what if we "explode" this ideal PC model to distance the user's input and output interfaces from the computing center to remote locations? (Refer to Figure 1) We end up with a view of the ideal PC's elements in a way that reveals the key characteristics that are of interest to Public Network engineers. The first characteristic one can see is that the information flow is uni-directional in nature rather than bi-directional. This is because the user's interaction with the computer uses separate and very different channels for input and output. The second characteristic is the asymmetrical nature of the bandwidth requirements. The user's input is narrowband in nature. On the other hand, the computer's audiovisual presentations are broadband in nature. Together, these characteristics offer an unique opportunity to network engineers that can view ITV networks as the integration of existing PSTN and cabled or wireless broadcast television networks.

To take advantage of this opportunity will require a paradigm shift in network architecture thinking. Conventionally, network engineers view the location of the network interface for the consumer as being between their PC and the external world. This is inappropriate for a Public Network system where "the network is the computer". We need to move the network interface to a point between the computer and the user's input devices. A telecomputer system has "two" network interface channels to the consumer. One is the narrowband "input network" interface channel to the computer and the other is the broadband "output network" interface channel to the consumer's TV. When viewed in this way, the computer in the PC telecomputer model can be centralized within the Public Network. Once the computer is taken out of the home and centralized, it is then possible to provide the consumer with economic access to the ideal PC. Centralizing the ideal PC computer offers enormous economic benefits to consumers and operators alike. One, consumers can share these ideal PCs based on traffic engineering rules. Two, the consumer is now relieved of the headaches and costs associated with PC maintenance and obsolescence. When centralized, the operator could then also provide the consumer with economic and transparent migration paths to more advanced microprocessors, software, graphics, intelligence, and faster network connections to the external world when, and only when, necessary or desired. Most important, by centralizing these ideal PCs, we simplify the task of network management and improve network reliability by eliminating the unnecessary complexity of data networking devices and a PC in every home. On the output side of the centralized ideal PC, the RGB signals that would normally go to the consumer's monitor can be converted to the NTSC broadcast video standard. Doing so offers the operator several network channel options for delivering the output interface to consumers. The most likely predominate form of delivery will be as just another CATV cable channel. But, unlike the existing telecomputing models for interactive TV, this approach will work as well with the over-the-air, satellite, wireless, and switched forms of realtime video delivery. By using the existing NTSC broadcast video standard as the output interface standard for ITV there will then exist the potential for multiple ITV network operators competing for their share of the ITV audience. If the user's input network channels are equally available, it will then be possible to offer competitive ITV programming universally. Fortunately, input network channels to a centralized ideal PC are already available via the copper pairs of the PSTN. Any back channel or upstream signalling and voice capability added to the CATV coaxial distribution plant would be redundant and a totally unnecessary economic burden on the NII. The existing PSTN is more than adequate to meet the narrowband bandwidth requirements of input network channels and is already universally available.

WHAT THEN SHOULD BE THE ROLE OF THE TSB ?

As the most proliferous element of a growing telecomputer system, the first thing the TV set-top box should not be is a camouflaged substitute for the consumer's ideal PC. A smart TSB cannot even come close to meeting the demanding requirements of the ideal PC. This misguided role for the TSB would add the largest and most unnecessary economic burden on the NII. Compromises in smart TSB design, be they in the power of it's processor or the amount of resident memory, will not lessen the economic burden. A waste of money is still a waste of money. Smart TSBs, as the primary focus for ITV applications, are destined to be failures in a mass media network and public utility environment. The Public Network will not benefit from an ITV architecture where a key network element is already obsolete when first deployed. Another reason the current and conventional TSB telecomputer model will fail is that it does not deliver a medium that lives up to the Public's expectations for interactive television. ITV must be more than selecting a movie, pizza, catalog shopping, and access to so-called second level gateways to other information sources. "Interactivity" is much more than just "selectivity". Interactive television, as a new mass medium, must deliver a completely new entertainment experience to consumers. Downloading adolescent arcade games or simple prompts for play-along TV game shows to the limited memory of smart TSB will just not cut it in the eyes of a mass market. Consumers are looking for and have been promised more. However, with the ideal PC, an embryonic form of the evolving interactive multimedia experience can be found on CD-ROM publications. Understandably, the inherent limitations of the TSB telecomputer model used in today's so-called ITV trials has failed to attract any interest from the artists, programmers, and creators of this new interactive multimedia art form. For a mass medium where, as they say, "content is king", this will not do. The focus of ITV program development must shift from the TSB to centralized ideal PCs within the Public Network. In this way, interactive programs developed for CD-ROM and the ideal PC are easily ported to the ITV platform as an additional distribution network and ancillary market for the mutual financial benefit of the parties involved. The TSB cannot be at the center of ITV, nor should ITV program development wait for or depend on it. The role for the TV set-top box is, as always, only as a mass media network broadcast receiver. Whether a network's broadcast signals reach the consumer directly from a satellite, CATV headend, over-the-air, or a telco central office, the TSB's role is still the same, It is a mass media network, primarily broadband, receiver. The Public Network does, however, need a next generation model of TSB that better recognizes a larger role for receivers in the evolving NII for the next several generation of Americans.

The most important NII element that this new TSB will offer the evolving Public Network is the potential for a common carrier form of video distribution. To realize this potential, the TSB and it's associated broadband distribution networks must support addressability and approach virtual VOD capacity. (Ref. 1) For ITV, the TSB is just an addressable broadcast media receiver presenting NTSC output network channel interfaces to consumers. However, should a consumer choose to interact with the ITV programming offered, as previously described, the PSTN already offers back-channel or upstream input network channels on a common carrier basis. As a result, the TSB role has no transmission requirements in the evolving NII. It is still only a mass media broadcast receiver.

OF MICE AND TV REMOTES

In the centralized ideal PC model for ITV, the computer delivers the audiovisual output interface to consumers as video signals over ordinary NTSC distribution networks. To visualize this, just watch any TV news broadcast where high-resolution graphics constantly appear over the shoulders of newscasters. What distinguishes ITV from passive TV, however, will be the opportunity for the consumer to interact with such computer generated images. Such TV graphics, when generated by an ITV program, will present the output interface to the consumer. The central issue for engineers then becomes the input interface for ITV networks. What is the consumer's point & click remote signalling unit? Will it be a mouse, a cable TV remote, or something else? Will the consumer's remote unit require a special data network device to transmit signalling? In the telecomputer model championed in this paper, initially, the consumer's point & click remote signalling unit will be the touchtone phone. Unlike computer mice and cable TV remotes, the telephone already has an ubiquitous upstream signalling network in place. This means the Public Network does not have to bear the economic burden of upgrading the cable networks for two-way capability in order to create this new interactive mass medium. The telephone is also the only potential ITV input device with the inherent capability to send voice input and commands to centralized computers without the aid of data network devices. We just need to look at the telephone's DTMF keypad in a different way. It was, after all, created to permit caller access to computers.
Point & Click. If any concept can be credited with opening the world of computers to the masses, it has to be "point & click" and the intuitive graphical user interface. Move a pointer on the screen to the desired selection and press a button. In entertainment software, press certain buttons to move animated characters to the right or left and others to trigger actions. Whether one is accessing a kid's game, PC, workstation, or supercomputers, it doesn't get any simpler than that. A set of buttons to control directional movement and a set to trigger actions. Certainly the twelve buttons of the touchtone phone can do that. We just need to define the DTMF signalling protocol.

"DTMF-TV", A NEW TELECOMPUTER MODEL

We must evaluate any new mass medium from the perspective of the consumer. For an interactive one, network engineers must focus on what buttons those consumers are being asked to push, how those buttons connect to and signal the network, and at what costs. Past attempts to create an interactive mass medium have always run smack into the same "chicken or the egg" problem. No mass audience, no compelling programming. No compelling programming, no reason for the mass market to purchase the necessary hardware and software required to interact with the computer programming. The "DTMF-TV" telecomputer model solves this classic dilemma. The mass market of today's TV audiences does not need any new hardware or software to access compelling interactive TV programming, just their existing TVs and touchtone telephones. To create the DTMF-TV telecomputer model, explode the ideal PC model to centralize the computer and divide it's I/O channels into separate user interface networks. For presenting an output interface to the consumer, convert the computer's RGB signals to the NTSC standard so it can then be delivered to the masses over any existing video network. The video network may be a free over-the-air broadcast, premium, or a pay-per-view (PPV) channel available only to cable subscribers. With NTSC video as the ITV output standard, it does not matter what type of video channel as long as the programming is delivered in realtime. What will matter, since this GUI may be presented to the regular and passive viewing TV audience, is that the multimedia programming be compelling enough to create a desire in the consumer to interact with it. Content is indeed, king. Assuming that the ITV programming is compelling, the output interface should indicate to the viewing audience that they can interact with the program by calling the displayed telephone number. By calling this number, the user will then be connected to the DTMF protocol converter sending input and control signals to the ITV program's computer. (Refer to Figure 2) This may be an 800, 900, long distance or local number. For most ITV programs it will likely be a local number since most telcos do not charge for incoming calls. The audiovisual output interface, possibly including voice prompts, should also inform the audience of potential users that the input interface will be their telephone keypad.
Central to the DTMF-TV telecomputer model is the DTMF protocol that will represent the user's ITV input interface to the computer. It is important that this interface be as intuitive to the user as GUIs strive to be on the output side. In developing this patented approach to interactive TV (Ref. 2), an analysis of various point & click devices yielded their common characteristics. Whether a mouse, joystick, trackball, or game controller; they all use a dual element set (DES) protocol. Signals are divided into a set for directional movement and a set for triggering actions. Considering the computer mouse as the simplest and most intuitive, it controls the directional movement of a pointer on the x, y, and diagonal axes and provides for triggering. Most mice include two, and some include three, buttons for triggering program actions. Applied to the standard 3 column by 4 row telephone keypad, the bottom row or "*", "0", and "#" keys appear to be the most appropriate keys to mimic the trigger buttons of most computer mice and game controllers. This leaves the top three rows where the outer ring of eight keys surrounding the "5" key as the intuitive choice for directional movement. This is easy to visualize. Looking at your telephone keypad, the ITV user will press "4" to move the pointer to the left, "6" to move it to the right, "2" for up, "8" for down, and so on. If presented consistently, it should not be difficult for the public to learn these basics of the DTMF-TV telecomputer input protocol. Standardized as the PhoneMouse(tm) ITV interface, the telephone keypad is capable of more complex actions by assigning special roles to particular keys. By assigning the "5", "0" and "#" keys to the role of "cyclers", this ITV interface adds even greater telecomputer functionality. For example, as the virtual "ALT" button, pressing the "5" key can instruct the ITV program to cycle through a list of alternative user control, computer, and network options. The user's remotely signalled selections from these alternatives may be represented on the screen as different pointer icons. Similarly, pressing the "#" and "0" buttons can remotely signal the ITV computer program to cycle through the ITV program's menu of main functions and sub functions, respectively. Other buttons will also have specific roles as PhoneMouse-based ITV programming develops. The bottom line is, when viewed in a new and unconventional way, using the telephone will be the fastest and most economical approach to interactive television.

CONCLUSIONS

The DTMF-TV telecomputer model offers the most economical approach to interactive TV. The DTMF-TV telecomputer model offers an approach that can provide for the most Public Network competition between ITV programmers; be they RBOCs, IXCs, MSOs, Broadcasters or other Electronic Publishers. The DTMF-TV telecomputer model is the only one that can provide the NII with equitable and affordable universal access to our evolving interactive multi-Media Information Superhighway. REFERENCES (1) - THE AWAKENING 2.0 (C) 1991 : The Comments of GNOSTECH Incorporated to the FCC's Proposed Rulemaking on Video Dialtone (Common Carrier Docket 87-266) (2) - United States Patent No. 5,236,199 : INTERACTIVE MEDIA SYSTEM AND TELECOMPUTING METHOD USING TELEPHONE KEYPAD SIGNALLING