From telecom@eecs.nwu.edu Sat Feb 23 14:32:56 1991 Received: from hub.eecs.nwu.edu by gaak.LCS.MIT.EDU via TCP with SMTP id AA14793; Sat, 23 Feb 91 14:32:48 EST Resent-Message-Id: <9102231932.AA14793@gaak.LCS.MIT.EDU> Received: from uunet.UU.NET by delta.eecs.nwu.edu id aa03837; 23 Feb 91 9:11 CST Received: from world.std.com by uunet.UU.NET with SMTP (5.61/UUNET-primary-gateway) id AA26971; Sat, 23 Feb 91 10:10:37 -0500 Received: by world.std.com (5.61+++/Spike-2.0) id AA26173; Sat, 23 Feb 91 10:10:25 -0500 Date: Sat, 23 Feb 91 10:10:25 -0500 From: Scott J Loftesness Message-Id: <9102231510.AA26173@world.std.com> To: telecom@eecs.nwu.edu Subject: Apple Petition (Text) Resent-Date: Sat, 23 Feb 91 10:43:58 CST Resent-From: telecom@eecs.nwu.edu Resent-To: ptownson@gaak.LCS.MIT.EDU Status: R [Note: the following transcription of Apple's Data-PCS Petition for Rulemaking is provided courtesy of Telecom, CompuServe's Telecommunications Issues Forum (GO TELECOM), and HamNet, CompuServe's Amateur Radio/Scanning/SWL Monitoring Forum (GO HAMNET). HamNet's Packet Radio section provides a wealth of resources on amateur packet radio technology, specifications, software, etc.] [Note: This petition has been assigned RM-7618 by the Federal Communications Commission. Comments should be sent to the Secretary, FCC, Washington, DC 20554 by March 7, 1991. Clearly indicate RM-7618 on your letter.] Before the FEDERAL COMMUNICATIONS COMMISSION Washington, DC 20554 In the Matter of Amendment of Section 2.106 of the Commission's Rules to Establish A New Radio Service for Local Area High Speed Data Communications Among Personal Computing Devices PETITION FOR RULEMAKING "DATA-PCS" David S. Nagel Vice President Advanced Technology Group Apple Computer, Inc. 20525 Mariani Avenue, M.X. 75-6J Cupertino, California 95014 (408) 974-5124 OF COUNSEL Henry Goldberg GOLDBERG & SPECTOR 1229 Nineteenth Street, N.W. Washington, D.C. 20036 (202) 429-4900 January 28, 1991 -------------------------------------------------------------------------- SUMMARY Apple Computer, Inc. ("Apple") proposes that the FCC initiate a rulemaking to allocate 40 MHz in the 1850-1990 MHz band to a new radio service to be used for high-speed, local area data communications service ("Data-PCS") between and among personal computers. FCC action is urgently needed because the computer industry is rapidly developing technologies to meet the requirements of computer users, that will be severely handicapped without Data-PCS. The development of computer technology over the past two decades has been characterized by two primary features: 1) computers have been networked, using cabling and common carrier facilities, to give users access to information from a variety of sources; and 2) there has been a steady movement towards placing the power of the computer directly into the hands of the user, wherever the user may be. As personal computer technology now moves from the desk-top to the briefcase, the networking and portability features will become mutually inconsistent unless the networking capability becomes as personal and portable as the computer itself. With such a networking capability, a person could communicate with his or her peers and could access files, peripherals, and the gateways of wired and wireless data networks, all within a "local area" of 50 meters. The development of Data-PCS, therefore, will facilitate spontaneous, collaborative computing in the work-place and in educational settings, thereby increasing the productivity and efficiency of people in these environments. There are, however, presently no technologies and no radio services that can be used to create the shared electronic space necessary for collaborative computing, principally because no existing technology or service can assure consistent, high- quality, high-capacity data communications in a spectrum-efficient manner. Apple, therefore, is proposing the creation of a new radio service to be devoted primarily and exclusively to local area, high speed data communications to support collaborative computing and spontaneous networking. As conceived by Apple, a Data-PCS radio service would: - be accessible to users of personal computers without imposition of licensing obligations, network connection fees, or air-time charges; - be open to any computer manufacturer's products and any network access and usage scheme that complies with regulatory requirements; - be regulated in a manner that assures non-discriminatory access to assigned frequencies by compatible devices for like purposes; and - have flexibility built into the initial regulatory scheme to encourage innovation in and the evolution of Data-PCS technologies and services. In particular, Apple urges the Commission to allocate 40 MHz between 1850 and 1990 MHz for Data-PCS, preferably 1850-1890 MHz, because these frequencies have optimum propagation characteristics for local area, in-building use. Such an allocation will allow several networks with data rates of, for example, 10 Mbps, to coexist in the same location. The Commission should model the basic regulatory structure for Data-PCS on Part 15 of the Rules, relying on manufacturers and the equipment authorization process rather than on individual licensing, to assure compliance with regulatory requirements. These regulatory requirements would: - confine Data-PCS transmissions to a maximum power limit of one watt of output power; - permit the use of antenna directionality, to take full advantage of reuse of frequencies; - require all users to transmit data in packetized form, with a required minimum period of "listening" for traffic before transmitting and a maximum permitted duration of continuous channel occupancy; - require manufacturers to disclose the channel usage and access schemes employed by all Data-PCS equipment, with the exception of encryption schemes, and not authorize systems using schemes intended to exclude fair access to the frequencies by others; - require manufacturers to assign each transmitter a universal ID to be transmitted in each transmission sequences; and - along with all PCS equipment and similar devices, protect the health and safety of users by including such devices under the general guidelines of GEN Docket No. 79-144. Finally, Apple requests that the FCC expedite consideration of this proposal and not defer action awaiting resolution of many complex questions associated with establishment of a regulatory framework for voice PCS. There is an immediate and critical need for Data-PCS, which cannot be met using any other frequencies or radio services. Unlike voice PCS, Data-PCS is not simply a more efficient, or less expensive, means of providing an existing service. The Data-PCS capability does not exist at present. It will require substantial investments to create Data- PCS, let alone to develop it to its full potential. Those investments cannot be made until the FCC allocates sufficient bandwidth and establishes the operating conditions that will enable Data-PCS to flourish. The urgent need for Data-PCS is underscored by international considerations facing the U.S. computer industry. At present the U.S. industry leads the world in personal computer technology, but it is being strongly challenged by Japanese companies. The U.S. industry, if it is to remain competitive, must be in the forefront of developing a wireless capability for personal computers. It can do so if the FCC creates an environment for Data-PCS now, in advance of WARC-92. Initiative by the FCC will encourage the spread of Data-PCS with a de facto U.S. standard and thus will substantially enhance the competitive posture of U.S. computer manufacturers in the world market. ---------------------------------------------------------------------------------------------------------------- TABLE OF CONTENTS SUMMARY TABLE OF CONTENTS PETITION FOR RULEMAKING I. INTRODUCTION II. DEVELOPMENTS IN PERSONAL COMPUTER TECHNOLOGY, REQUIREMENTS FOR COLLABORATIVE COMPUTING, AND NATIONAL INTEREST GOALS CREATE AN IMMEDIATE AND URGENT NEED FOR DATA-PCS A. Personal Computer Technology is Moving in the Direction of Smaller, More Powerful Computers B. Increased Productivity and Efficiency in the Work Place and the Classroom Require a Capability for Spontaneous, Collaborative Computing C. The United States' National Interest Will Be Served by Development of Data- PCS III. NO EXISTING TECHNOLOGY OR RADIO SERVICE, AND NO PROPOSED PERSONAL COMMUNICATIONS SERVICE MEETS PRESENT AND ANTICIPATED NEEDS FOR DATA-PCS A. ISM Bands Do Not Provide a Suitable Environment for Data-PCS B. Data-PCS Cannot Be Provided as Part of Proposed Voice PCS and Cannot Share Frequencies with Voice PCS IV. THE FCC SHOULD CREATE A NEW DATA-PCS RADIO SERVICE WITH 40 MHZ OF BANDWIDTH A. The FCC Should Allocate 1850-1890 MHz for the Exclusive Use of Data-PCS 1. Frequencies at 1850-1990 MHz offer optimum propagation for Data-PCS 2. Forty MHz is the minimum amount of bandwidth that should be allocated at present B. The Regulatory Framework for Data-PCS Should Be Modeled on Part 15 and Should Assure Spectrum Efficiency, Compatibility of Usage, and Non- Discriminatory Access to Assigned Frequencies 1. Spectrum Efficiency is Best Assured by Adoption of Reasonable Restrictions on the Output Power of Data-PCS Transmitters 2. Compatibility of Users and Nondiscriminatory Access to Frequencies Are Best Assured by Requiring that All Users Transmit Data in Packetized Form a. Disclosure by Manufacturers b. Requirements for a universal identification c. Non-disclosure of encryption schemes C. The Commission Should Adopt Health and Safety Standards for all PCS Equipment V. CONCLUSION AND REQUEST FOR EXPEDITED CONSIDERATION ---------------------------------------------------------------------------------------------------------------- PETITION FOR RULEMAKING Apple Computer, Inc. ("Apple"), pursuant to Section 1.401(a) of the Federal Communications Commission's ("FCC" or "Commission") Rules and Regulations, 47 C.F.R. Section 1.401(a) 1988, respectfully requests that the FCC initiate a rulemaking to allocate 40 MHz in the 1850-1990 MHz band to a new radio service to be used for high-speed, local area data communications services ("Data-PCS") between and among personal computers produced by any manufacturer. Such a radio service will encourage individual productivity and the "autonomous creation of knowledge" /1/ by facilitating ad hoc local networking of individuals among their peers and with their file servers and other peripheral resources, without requiring pre-configured wired networks. The present lack of suitable frequencies for wireless, local area networking transmissions is retarding the full development of personal computer technology and could have a severe adverse effect on the future position of the United States in the worldwide services-based economy. Apple is asking the FCC to begin the process for creating the local area, wireless dimension of a national information infrastructure that will carry information to people any time and any place. The first, and most important, link in the information infrastructure is the "local area" (i.e., 50 meters) in which a person must communicate with his or her peers, files, and peripherals. To qualify as an effective infrastructure, "a resource must be widely available, easy to use, and inexpensive."/2/ This local area wireless infrastructure does not exist at present, and it cannot be created unless and until the FCC dedicates sufficient radio spectrum for this purpose. FCC action is urgently needed because computer users are requiring capabilities, and the computer industry is rapidly developing technologies to meet the needs of computer users, that will be severely handicapped without Data-PCS. Moreover, the United States must act quickly in dedicating frequencies for Data-PCS in order to establish a presence in advance of WARC-92 and to set a de facto worldwide standard for what is becoming the most rapidly growing segment of the personal computing marketplace. Finally, as set forth below, Data-PCS will serve the national interest by fostering the productivity of the United States' economy and enhancing its international leadership in personal computer technologies. I. INTRODUCTION Apple Computer, Inc., develops and manufactures personal computers. Since its incorporation in 1977, Apple has grown to a publicly-held, Fortune 100 company with more than twelve thousand employees. Sales in fiscal 1990 exceeded $5.5 billion, with sales outside of the United States accounting for approximately 42 percent of Apple's revenues in 1990./3/ Apple's corporate philosophy is based on its conviction that personal computers can dramatically improve the way people work and learn, giving them new tools for doing both. We are moving from an industrial economy to an information-based, services economy. Innovations that enhance access to and use of information directly contribute to the national welfare. In the work place as well as in educational settings, the skills that people need most are information-processing skills and the resource people need most is access to information. To serve these needs, computers must have an open and capacious ability to communicate. Local area networks allow users to communicate quickly and inexpensively with computers attached to the network and to share information and resources. They also allow people to use their computers to form, dissolve and reform working groups in real time. For this reason, Apple is an industry leader in computer communications, with more than 70 percent of all Macintosh computers networked and more than two million AppleTalk(tm) local area network (LAN) nodes installed worldwide. Apple's range of networking and communications tools make the Macintosh computers the easiest-to-integrate computers on the market today, compatible with every major industry standard - Digital Equipment, IBM, Open Systems Interconnection (OSI), and Transmission Control Protocol/Internet Protocol (TCP/IP) environments. Apple's emphasis on interconnectivity among personal computers is not unique in the computer industry. The creation of personal computer networks is one of the fastest growing phenomena in the United States' corporate sector. It is estimate that, in 1990, U.S. companies connected 3.8 million personal computers in local area networks, an increase of 48 percent over 1989./4/ Internationally, worldwide revenues for PC local area network products grew at a compound rate of 46.3 percent between 1985 and 1988./5/ LAN technologies and their ability to enhance productivity are, however, limited today by the need for physical (wire) connections. As personal computers become more portable -- now shrinking to notebook size -- users have more compelling reasons to take their computers wherever they go, increasing their creativity and their productivity. Without wireless connectivity, users will be constrained in accessing the resources they need. II. DEVELOPMENTS IN PERSONAL COMPUTER TECHNOLOGY, REQUIREMENTS FOR COLLABORATIVE COMPUTING, AND NATIONAL INTEREST GOALS CREATE AN IMMEDIATE AND URGENT NEED FOR DATA-PCS A. Personal Computer Technology is Moving in the Direction of Smaller, More Powerful Computers Since the early days of computer rooms, computer development has moved steadily towards placing the power of computer technology directly into the hands of the user, wherever the user may be. As the technology now moves from the desk-top to the briefcase, the communications capability must become personal and portable. The new, smaller personal computers are the fastest growing segment of the personal computer market. Sales of laptops and notebook-sized computers grew 58.6 percent in 1990, compared to growth of the total personal computer market of just 5.3 percent./6/ Sales in 1991 are expected to increase an additional 40 percent to approximately $3.9 billion./7/ Many forecasters expect sales to continue to grow by 40 percent to 40 percent a year and to account for at least one-third of U.S. personal computer sales by 1993./8/ It also has been estimated that nearly 40 percent of worldwide personal computer sales will be portable machines, as compared with only 14 percent now./9/ This growth in the sales of portable PC's has significant implications for the future international role of the United States' computer industry. United States manufacturers held a 61 percent share of world sales of computer systems in 1989, but this share is down from 81 percent in 1983. The Japanese share nearly tripled from 8 percent to 22 percent in the same period./10/ The future of the U.S. personal computer industry may well be determined by sales of portable PC's, a product category aggressively targeted by Japanese makers. Although Japanese companies have only a 9 percent overall share of the U.S. personal computer market,/11/ they produced an estimated 40 percent of the portable computers sold in the United States in 1990./12/ If it is to compete successfully in the next wave of PC technology, the U.S. computer industry must be allowed to develop its lead in wireless networking capabilities, to allow PC users to communicate spontaneously and in real time without regard to pre-wired network configurations. B. Increased Productivity and Efficiency in the Work Place and the Classroom Require a Capability for Spontaneous, Collaborative Computing Today's information-driven society is rarely static and requires the ability to start up quickly, relocate frequently, and respond to new needs. Wireless networks can offer immediate startup and re-configuration. They can, for example, provide communications - from one computer user to another in a work group, - among teachers or conference leaders with assembled scholars, - between a user and his or her printer or file server, or - from a scientist to a network "trail-head" that connects him or her to a national data network. The value of wireless computer technologies will be profoundly realized in educational settings./13/ Despite millions of computers now used in schools around the world, the true power of the computer as an aid to learning has been only partially realized. By the time today's five-year-olds complete high school, they will have to cope with thirty times more information than exists today. Preparing children and adults for the information age is a daunting challenge, but one that cen be met by taking full advantage of information and telecommunications technology. In today's budgetary environment, installing the cabling required for students to use computers collaboratively is prohibitively expensive. Even if funds could be found for hard wiring, it is excessively time-consuming and constraining to re-arrange a wired network, once installed, as teachers move students among reading, math, and writing groups. Data-PCS can make it possible for schools not only to afford to employ networked computers, but also to take advantage of "situation-driven" learning, that happens spontaneously and results from interactions among students and teachers. In the work place, as in the classroom, spontaneous interaction and collaborative processes can increase one's productivity. As one observer has expressed it recently: We need to build tools, technologies and network environments that encourage productive relationships, not just productive individuals... Unfortunately, most networks are designed with the goal of exchanging mail and messages rather than creating shared space where people can collaborate./14/ C. The United States' National Interest Will Be Served by Development of Data- PCS In addition to the inherent public-interest benefits of Data-PCS, creation of such a new radio service will also serve the national economic interest of the United States in two ways: first, by contributing to the increased education and productivity of our work force, and second, by helping maintain the leading role of the U.S. computer industry. Both benefits are well understood by the Commission. As Chairman Sikes recently stated: Mobile services are a proven means of boosting productivity in a services-based economy./15/ While Chairman Sikes was speaking of land mobile services in general, mobility at any scale provides flexibility, adaptability, and creative human interfaces that support innovation. Quoting former Secretary of Commerce Baldridge, Chairman Sikes noted that: Technology is America's competitive edge...and, it was our demonstrated capacity to "push the technological envelope" and then deploy the results in the form of new products which historically conferred a decisive commercial advantage./16/ The personal computer industry still has a uniquely American flavor. PC's, and the microchips that made them possible, are home-grown technologies. American computer technology emerges from American developers to meet American needs and tastes. The customer base in the United States is the largest in the world, and the United States is a net exported of hardware based upon or including computer technologies. As noted above, this U.S. leadership is being challenged by Japan's development of portable computer technology. The freedom to continue to innovate both computer hardware and ways of using computers is fundamental to maintaining the U.S.'s leading edge. Adding a wireless, local area communications capability is a next requisite step in the evolution of personal computers, and the U.S. industry must be in the forefront. To achieve this, Data-PCS must be created and developed in the United States. This alone will encourage the spread of Data-PCS in a U.S. configuration and will directly support the ability of U.S. companies to compete in the world market. The best way to establish a world standard is to get there first. In the discussion that follows, Apple demonstrates that Data-PCS cannot be accommodated within presently available technologies or frequency bands or within the services that have been proposed by PCN/PCS advocates. Apple then describes the proposed technical and regulatory characteristics of a new radio service to meet the unfilled and urgent need for Data-PCS. III. NO EXISTING TECHNOLOGY OR RADIO SERVICE, AND NO PROPOSED PERSONAL COMMUNICATIONS SERVICE MEETS PRESENT AND ANTICIPATED NEEDS FOR DATA-PCS Given the pressing need for a wireless, local area communications capability for personal computers, Apple and other computer companies have analyzed the existing technologies and radio services to determine if they can accommodate Data-PCS. In all cases, the answer has been no, principally because no existing technology or service can assure high-quality data communications in a spectrum-efficient manner. Digital data communications require a very high quality transmission environment. Information must be conveyed in such a way that it can be received with virtually total accuracy, no matter what happens in the transmission medium or in the communications equipment. Voice communications generally succeed even with errors in the transmission path of up to one percent (and error rate of 10-2), depending upon the nature and distribution of those errors. Digital data may be handled many times in a network; an error rate of 10-8 (one million times better than for voice) is a common minimum requirement, and still greater integrity is desired for some applications. In addition to that requirement for accuracy, effective data networks must provide for high-capacity data throughput, a capability for real-time information-intensive collaborative computing, and the ability to create spontaneous networks. These requirements cannot be met by existing technologies or services. For example: - Data modems used with the cellular telephone networks provide virtually nationwide coverage, but they offer very low data rates at high costs per unit of connection time. Some proposed digital cellular systems could allow for increased data rates, but air-time charges would preclude widespread usage./17/ - Several SMR-based mobile data networks provide metropolitan area coverage for an air-time fee and with limited data transmission rates. These networks are useful for many applications and can complement Data-PCS, but cannot substitute for it. - New personal communications networks offer or promise wide area -- even worldwide -- wireless data communications. Virtually all of them, however, would offer data communications only as a supplement to voice communications. These proposed networks are characterized by slow data rates and by airtime-based usage charges, both of which undercut these networks' utility in meeting the present and growing needs of personal computer users. - Motorola recently announced that it is acquiring licenses to implement its 18 GHz "WIN" in-building LAN. WIN-type systems will be complementary to Data-PCS, but transmission in the 18 GHz band may not penetrate partitions in many buildings and are, therefore, unsuitable for the full functionality desired for Data-PCS as it has been defined. Moreover, the requirement for frequency coordination and fixed hardware limits flexibility. - Infrared provides an often under-appreciated medium that satisfies many in-room computer network scenarios. Infrared does not, however, provide the mobile connectivity range desired for Data-PCS. - Various radio frequency picocell models, some operating as high as the oxygen absorption frequencies (60 GHz), have been proposed in Europe and elsewhere./18/ When these technologies become practical, they can offer efficient spatial reuse of frequencies but they cannot provide the coverage range needed for Data-PCS. Therefore, like other technologies, such high frequency systems can complement Data-PCS but cannot substitute for it. Faced with the limitations of existing and proposed services, Apple and other computer manufacturers began to experiment with and develop wireless computer products for the ISM frequencies, particularly using spread spectrum modulation techniques. Such use of the ISM frequencies had the obvious appeal of availability, with a regulatory structure that put the compliance burden on manufacturers and equipment suppliers and did not require individual licensing of computer users. It also had the considerable advantage of allowing consumer usage without the barrier of air-time or network connection charges. Despite these advantages, as discussed fully below, attaining adequate data communications services in the ISM bands has proven to be unfeasible. A. ISM Bands Do Not Provide a Suitable Environment for Data-PCS In 1985 /19/, the Commission opened the industrial, scientific and medical ("ISM") bands /20/ for unlicensed operation of devices which comply with the technical regulations of 15.247./21/ Section 15.247 permits spread spectrum modulation, allowing a maximum transmitter power of one watt. In 1990, /22/ the Commission modified the Rules to permit wider bandwidths when frequency hopping spread spectrum is employed. Apple supported that modification in the expectation that using spread spectrum modulation techniques in the ISM bands, pursuant to Part 15, would provide adequate spectrum resources to support Data-PCS./23/ After extensive testing, Apple has concluded that such operation in the ISM bands ultimately will be unworkable, because there is a strong likelihood of unpredictable, and essentially uncontrollable, interference in the ISM bands./24/ Data-PCS can share frequencies with users who operate under the same defined, predictable methods of channel access and traffic usage, but this orderly usage does not exist in the ISM bands. The present Rules permit a disparate group of users to access ISM frequencies for a wide variety of communications, and even non-communications, purposes. The very variety of uses, modulation schemes and power levels creates in the ISM bands the antithesis of the homogeneous and stable operating environment needed for data communications. In particular, the Rules permit relatively high-powered transmitters which can create uncontrolled interference to communications services sharing those bands, and do not offer sufficient latitude for users to employ techniques to deal with such interference./25/ The permitted levels of interference thus preclude the possibility of mutual accommodation. The presence of uncontrollable interference dominates the effort to achieve adequate throughput rates of data transmission on the ISM bands. As discussed above, digital data communications require a much higher quality transmission environment than voice communications. Digital communications networks can and do employ powerful techniques of anticipating and/or correcting errors introduced by the communications medium. These techniques themselves, however, increase the amount of information that must be communicated, and thereby increase the channel occupancy or the required bandwidth. In planning spectrum usage, it is necessary to strike a judicious balance between providing a high quality transmission environment and burdening the channel with the additional overhead required to assure robust data transfer. The more that users employ countermeasures against expected interference (e.g., higher power, high-performance antennas, more profound error correcting codes, and especially repeating transmissions in order to "get through"), the more probable it is that the result will be a ratcheting upwards into spectrum gridlock. This ratcheting is the antithesis of spectrum efficiency./26/ Accordingly, given present and anticipated operating conditions in the ISM bands, it would be reckless for the computer industry to ignore these trends and expect ISM frequencies to be a realistic medium for Data-PCS operation through the decade./27/ B. Data-PCS Cannot Be Provided as Part of Proposed Voice PCS and Cannot Share Frequencies with Voice PCS Apple has also examined the possibility of Data-PCS being provided as part of any number of voice personal communications services, or co-existing in the same frequencies with future voice PCS systems such as have been proposed by a wide variety of entities. We have concluded that it cannot. In the various PCS proposals that have been submitted to the FCC, data services would be provided along with voice, but would be subsidiary to voice services. These predominantly voice systems cannot be used for Data-PCS because they have been conceived and proposed in the context of today's telephone modems, which have only recently achieved useful data transmission rates in the range of 9,600 to 19,200 bps. This is far less than the data transmission rates necessary to provide significant computer data connectivity consistent with the needs of today's users and the abilities of computers to acquire and manipulate data. Apple next considered whether Data-PCS could share the same frequencies as proposed for voice PCS, but has found that Data-PCS could not coexist with voice PCS for reasons largely similar to those that preclude coexistence with other users in the ISM bands. Although voice PCS frequency bands would not present the same highly diverse mix of users as the ISM bands, the transmission and duty-cycle characteristics of voice PCS users are sufficiently different from, and indeed are technically antagonistic to, those of Data-PCS users as to preclude the possibility of sharing frequencies. Data-PCS cannot necessarily survive the potential presence, on the same frequencies, of one or dozens of voice PCS devices as described in current PCN proposals. Those devices, some of which may in fact be within arm's length of the computer, radiate enough RF power with continuous or long-duty cycle transmissions to communicate reliably with a PCN microcell base station that may be hundreds of meters away. Some data and voice transmission schemes cannot coexist simply because data transmission requires a higher signal-to-noise ratio than voice transmission to achieve the requisite low error rates. Attaining that higher signal-to-noise ratio for a wider-bandwidth signal necessary to accommodate effective high speed data transmission and prevail over voice signals requires more RF energy than may be feasible for portable, battery-powered devices or desirable from the viewpoints of intrinsic safety and of frequency reuse. A multiple access (i.e., shared) data network can most efficiently access a channel only when it needs to and, even then, only when it can sense that the channel is available; real-time voice transmission, however, generally requires one form or another of guaranteed channel time and bandwidth. Any protocol or channel access scheme that dynamically provides the necessary guarantees for real-time voice would, by definition, force any other service into accepting what is left. As the need for high-capacity data transmission increases, what is left over will not be enough. IV. THE FCC SHOULD CREATE A NEW DATA-PCS RADIO SERVICE WITH 40 MHZ OF BANDWIDTH Data-PCS is a new technology that will provide a completely new service. It is not an extension of or supplement to mobile or portable telephone services. As a new, stand-alone technology, Data-PCS needs a suitable operating environment that ensures robust, high speed, data communications on frequencies reserved exclusively for its use. The need for a user-driven, open-access high-capacity computer communications technology to support collaborative computer and spontaneous networking dictates the shape of Apple's proposal. - A Data-PCS radio service must be open to any computer manufacturer's products and any network access and usage scheme that complies with the regulatory requirements discussed below. - It must not impose licensing obligations or air-time charges upon users of personal computers. - It must have adequate bandwidth to support high-speed, highly-reliable data communications between and among various types of PC's and peripherals over a 50-meter range, primarily within a single building. - A Data-PCS radio service must be regulated in a manner that assures that the assigned frequencies will be used by compatible devices for like purposes and that there will be fair access to the frequencies for such devices and purposes. - Flexibility must be built into the initial regulatory scheme to encourage innovation in and the evolution of Data-PCS technologies and services. Apple's proposal meets these objectives and is described in detail below. A. The FCC Should Allocate 1850-1890 MHz for the Exclusive Use of Data-PCS Apple requests that the Commission create an exclusive allocation of 40 MHz for Data-PCS in the 1850-1990 MHz portion of the spectrum, preferably 1850-1890 MHz. 1. Frequencies at 1850-1990 MHz offer optimum propagation for Data-PCS Apple favors allocation of frequencies in the 1850-1990 MHz range primarily because of the propagation characteristics of these frequencies. Computers equipped for Data-PCS are intended to communicate within a radius of 50 meters on a single floor inside a building. This minimum coverage could encompass groups of people who work together and who share computer and peripheral resources, and who thus need to be networked together./28/ (At the same time, we recognize that frequency reuse is of profound importance in achieving efficient spectrum utilization.) Radio propagation within this radius will depend upon many factors, including building construction techniques, intervening walls and partitions, and the frequency of the transmission. Attenuation, fading, impulse response and multipath issues will affect the range and reliability, particularly as the radio frequency increases. Apple has performed tests in a variety of buildings, using frequency ranges of 900, 1990, 2400 and 5800 MHz and employing several modulation schemes. We have also examined published reports which generally confirm our experiments. We believe that, present frequency allocations and interference notwithstanding, the frequency domain of approximately 1600-2400 MHz provides a reasonable balance between effective penetration of in-building partitions and frequency reuse indoors, at RF power levels appropriate for small, portable computers equipped with miniature, non-optimum antennas. In addition to favoring frequencies in the 1850-1990 MHz range because of their propagation characteristics, Apple believes that they are desirable because regulatory authorities in other nations seem to be at the point of selecting frequencies in this range for other personal communications services -- a consideration that could enhance a worldwide overlay of Data-PCS within other personal communications scenarios./29/ 2. Forty MHz is the minimum amount of bandwidth that should be allocated at present While designation of any particular amount of bandwidth to be devoted to Data-PCS may be arbitrary, Apple believes that 40 MHz will be necessary and may adequately meet the present and foreseeable needs to be served by Data-PCS./30/ Current generations of personal computers typically communicate at speeds of several Mbps./31/ These rates are indicative of the bandwidth a Data-PCS channel needs to provide. Indeed, more than a single 10 MHz channel may be required in a particular location, since many environments have densities of computer users (or needs for multiple networks) whose network rate requirements can, in the aggregate, exceed the capacity of one such channel at any given time./32/ An allocation of 40 MHz is sufficient to permit several Data-PCS networks operating at rates of up to 10 Mbps, for example, to coexist in the same geographic area today, as well as to motivate technological innovations that can lead to higher data rates in the future. There is a clear relationship between the ability of a computer to process information and the communications bandwidth required to support the computer's capabilities. Progress in video and data compression and processing is currently enhancing wired networks. Similarly, wireless networks must also keep pace to support such innovative applications as multimedia combinations of text, images and sounds. In any event, recognizing that the amount of bandwidth that Apple is request could ultimately prove limiting, this Petition seeks to meet as many needs as can be addressed practicably in the present regulatory context. In this regard, while an immediate allocation of a contiguous 40 MHz would enhance the potential of developing technologies allowing substantially higher data rates, if such an allocation is not immediately available, Apple requests allocation of 40 MHz divided into at least 10 MHz portions, for example, 1850-1860 MHz, 1910-1930 MHz, and 1980-1990 MHz./33/ Moreover, the allocation could be implemented over a period of eight years, if at least 10 MHz are allocated at the outset, the remainder are identified, and there is an assurance that the balance will be allocated in due course. For example, the phased release of the 40 MHz could proceed as follows: the first 10 MHz should be released for Data-PCS by the end of 1991 and cleared of interference sources no later than the end of 1992, with further 10 MHz increments similarly released and cleared at two-year intervals thereafter, over the balance of the decade. Phased release of frequencies also could allow existing users of the frequencies to re-tune their radios within the band, rather than shift to other bands immediately, thereby reducing the costs of transition. Ad-hoc industry agreements providing for compensation to existing users could expedite the process. Furthermore, the FCC's rules could evolve with the release of frequencies. The regulations that can be formulated now are unlikely to reflect all the possible technical changes that will take place during the decade. A phased release of frequencies would give industry the opportunity to advise the Commission on any appropriate changes in the rules. B. The Regulatory Framework for Data-PCS Should Be Modeled on Part 15 and Should Assure Spectrum Efficiency, Compatibility of Usage, and Non- Discriminatory Access to Assigned Frequencies Determining the regulatory requirements for Data-PCS is a complex undertaking. At this juncture, Apple proposes only general objectives and a regulatory framework for meeting them, and does not propose precise specifications to be incorporated in FCC rules for Data-PCS. There should be a thorough dialogue, within the industry and between the Commission and industry, both within and without the rule-making process to refine the details. At this early stage in the development of Data-PCS, the FCC should set broad standards to assure that the allocated frequencies actually will be used for the purposes that justify allocating the frequencies to Data-PCS in the first place. These standards should be flexible enough to encourage innovation and technological evolution, but not so broad as to allow a variegated mix of users with disparate and, possibly, incompatible technical operating characteristics. As a sine qua non, the FCC should model the Data-PCS regulatory scheme on Part 15 of the Rules. Rather than licensing common carriers or other types of service providers to create Data-PCS networks and provide service to users, public interest goals best can be achieve in a Part 15-type framework, with software-denominated "networks" and by relying on manufacturers and the equipment authorization process to assure compliance with regulatory requirements. Similarly, following a Part 15 model, there should be no licensing requirement for individual users. Use of the spectrum should be open and "transparent" to the user, much as radio listeners, broadcast TV viewers, and cordless telephone users freely avail themselves of those spectrum-based services without air-time charges or network connection fees imposed by service providers. If it is deemed necessary to impose a license fee for spectrum usage, as discussed below, it can be imposed pursuant to the FCC's fee Rules and collected and administered by the equipment manufacturers. Above and beyond the Part 15-type regulatory approach, Apple proposes that the FCC adopt operating requirements, as well as equipment authorization standards, that serve to assure: - spectrum efficiency - a technically compatible mix of users; and - nondiscriminatory access to the frequencies. 1. Spectrum Efficiency is Best Assured by Adoption of Reasonable Restrictions on the Output Power of Data-PCS Transmitters Spectrum efficiency goals will be served by confining Data-PCS transmissions to a maximum output power of one watt. One watt of RF power into a compact, portable antenna at 1900 MHz is the minimum power that, inside many buildings, supports robust through-the-wall data communications throughout a 50-meter radius. This power level would enable micro-cellular frequency reuse in larger premises, but would preclude Data-PCS from being used primarily for point-to-point services that could be more appropriately located in other radio services or by wired networks through gateways. Since Data-PCS connectivity can in many environments be achieve with less that one watt of RF power, Apple proposes that equipment capable of more than ten milliwatts of power should be required to have provisions for adaptive power control to utilize the lowest power level that meets network connectivity requirements. In suggesting power limits, Apple favors a simple expression of output power, rather than regulation based upon power spectral density, which might indirectly inhibit technical latitude in developing modulation schemes for providing Data-PCS./34/ In addition, the rules should encourage the use of directionality, where practical, to take full advantage of geographic discrimination to reduce the potential for interference. Apple, therefore, suggests that for each 2 dBi of antenna gain, or portion thereof, in excess of 6 dBi, a 1 dBi reduction (only) in maximum terminal power output should be required./35/ 2. Compatibility of Users and Nondiscriminatory Access to Frequencies Are Best Assured by Requiring that All Users Transmit Data in Packetized Form To assure that there will be a technically compatible mix of users and that all users have nondiscriminatory access to the assigned spectrum, the FCC should require that all users transmit data in packetized form./36/ Rather than adopt detailed regulations in this respect, the Commission should rely heavily on the cooperative efforts of the computer industry and of private sector standards organizations. In particular, the Commission should be guided by the IEEE 802.3 standard describing Carrier Sense Multiple Access (CSMA), a means by which two or more stations share a common medium. At the simplest level, a station listens on a channel it desires to use and waits until the channel is available before initiating transmission. The information to be transmitted is divided, if necessary, into short packets, which may include both data and network-control symbols. The duration of any such transmission packet, and thus of continuous channel occupancy, is limited./37/ Packetizing of digital information is the fundamental basis for AppleTalk(tm), Ethernet and virtually all other computer networks. Packetization and channel access are generally independent of modulation scheme and can be employed with spread spectrum as well as non-spread and channelized operational models, and can be applied to CDMA, TDMA, FDMA and combinations thereof. Because it supports powerful means for preventing or correcting errors, packetization uses the spectrum very efficiently. Accordingly, Apple proposes that the minimum specification needed to control channel access and usage are: - A required minimum period of sensing the channel for activity before considering that the channel is available for transmitting. If activity is sensed, transmission is deferred until the channel is free. - A maximum permitted duration of continuous channel occupancy. Except for these minimum requirements, the Commission should not dictate or restrict the architecture of Data-PCS networks. In particular, it would not be desirable to have regulations that would require a centralized network management resource or a pre-existing superstructure, any more than there should be regulations to prevent them./38/ a. Disclosure by Manufacturers The Commission should, as part of the equipment authorization process, require disclosure by the manufacturer of the channel usage and access schemes employed by that equipment for Data-PCS. We believe that modulation schemes, error correction provisions, channel access protocols and provisions for fairness of access, interference rejection schemes, receiver and transmitter emissions and reception masks, and other pertinent characteristics that define channel access and occupancy should be fully revealed. In this way, industry participants can work in a known environment. Such disclosure can expedite de facto standardization of ways to share the RF medium; inefficient schemes can be supplanted by better ones; and responsible use of the spectrum will be encouraged. The Commission should not authorized systems employing schemes specifically intended to limit or exclude fair access by others. b. Requirements for a universal identification To provide further incentive and means for the computer industry to agree on standards and protocols regarding equipment for and usage of Data-PCS frequencies, the FCC also should require manufacturers to assign each transmitter a universal identification ("ID") that is transmitted, in whole or sufficient part, during each transmission sequence. Industry groups and standards organizations will cooperate with the Commission in selecting the proper form of this ID. Moreover, if the Commission wishes, such a universal ID assignment scheme could provide a means for manufacturers to collect a fee for spectrum usage from PC users and remit such fees to the Government. Such fees have proposed for other services and are highly controversial. There is, however, substantial precedent for computer users to pay fees for intellectual and other intangible properties, such as software and software updates. c. Non-disclosure of encryption schemes The only exception from the disclosure obligation should be for encryption schemes. Data transmitted over wireless media are inherently less secure than data transmitted on wired systems. The Commission should encourage or require manufacturers to provide means to protect data from interception and misuse. In this case, manufacturers should have the right to keep such encryption schemes proprietary. C. The Commission Should Adopt Health and Safety Standards for all PCS Equipment Finally, every effort should be made to insure and promote the intrinsic safety of products used by consumers in the work-place and in schools. The Commission should include Data-PCS products, as well as other technologies such as voice PCS, under the guidelines and requirements of GEN Docket No. 79-144./39/ Equipment authorization applications should indicate clearly the means taken to insure that the RF radiation of a particular device, used under the conditions reasonably expected to prevail, will not have a significant environmental impact. For example, reference to ANSI C95.1-1982 and showing of compliance might be required until the Commission adopts another Protection Guide./40/ V. CONCLUSION AND REQUEST FOR EXPEDITED CONSIDERATION More than seven years ago, another U.S. computer company /41/ filed comments in an FCC proceeding /42/ in which the Commission proposed to allocate frequencies in the 900 MHz band for a new low power personal communications service. The comments pointed to a growing need for a wireless communications capability for personal computers and stated that: The rapidly-expanding personal computer market is expected to create new requirements for communications between computers. This need has been met to date by either connecting computers to common carrier wire lines, or interconnecting adjacent units via a local area network. An evolution to smaller and personal units, however, will bring pressures upon more mobile and flexible communications techniques. A full solution to this potentially very large future service requirement should involve radio communications which use state-of-the-art computer communications techniques, including packet radio, local area networking and packet networking. No allocation was made in the 900 MHz frequencies at that time and computer companies have been unable to make other provisions for meeting this "potentially very large service requirement." Apple and other companies carefully examined use of the ISM frequencies, particularly in conjunction with spread spectrum modulation techniques, to meet this service requirement. As shown above, however, ISM frequencies do not offer a viable environment for data communications even in the foreseeable future. Even though the urgency of the need is causing some companies to continue to develop products for the ISM bands, the risks of interference-plauged operation in those bands are too great to make the substantial investments that are necessary to develop Data-PCS to its full potential. Those investments will be made, and the manifold benefits of Data-PCS will be secured, only if the FCC allocates sufficient bandwidth for this purpose and creates, by its regulations, the operating conditions that will enable Data-PCS to flourish. The need for expedited consideration of Apple's Petition is dictated by the requirement to make the investments now for the long-term development of personal computer products using Data-PCS. The need for Data-PCS was urgent when it was expressed more than seven years ago. It has become critical now, all the more so because of the imminent decisions that will be made in the context of WARC-92. Accordingly, a decision with regard to spectrum allocations for Data-PCS cannot await resolution of the complex range of issues associated with development of a wide variety of voice PCS technologies. The decision is overdue now. Apple's chief executive officer, John Sculley, recently stated to an educational group that: The key strength of twenty-first century organizations will be not their size or structure, but their ability to simultaneously unleash and coordinate the creative contributions of many individuals./43/ Data-PCS is one of the tools that will enable individuals to realize this vision. By taking the lead to create a Data-PCS, the FCC will be taking an essential step to assure that organizations in the United States -- both educational and commercial -- will be empowered to compete in the twenty first century and that the United States computer industry will have the versatility and strength to continue its contributions to our economy and to our society. Respectfully submitted, Apple Computer, Inc. David S. Nagel Vice President Advanced Technology Group Footnotes: 1 Harry Tennant, Technology 2001: The Future of Computing and Communications, MIT Press, 1990. 2 Michael L. Dertouzos, "Building the Information Marketplace," MIT Technology Review, January 1991, p.30. 3 October 1989 through September 1990, as reported in Apple's 1990 annual report. 4 Business Week, "Taming the Wild Network," October 8, 1990, p.143. 5 Financial Times, "Glue for the Global Village," November 20, 1990, Section IV, p.1. 6 Wall Street Journal, "PC Industry Pins Big Hopes on Laptops," October 29, 1990, Section B, p.1. 7 Id. 8 Wall Street Journal, December 24, 1990, p.12. 9 New York Times, "Japanese Portables Threaten American Lean in Computers," November 24, 1990, p.1. 10 Id. 11 Id. 12 The present Japanese foothold in portable machines could also presage the outcome of the U.S. and Japanese computer industries' battle for the European computer market, which totalled $27.8 billion in overall sales in 1989. Europe's present estimated installed based of 19 million PC's is expected to grow to 43 million by 1994. Financial Times, "Glue for the Global Village," November 20, 1990, Section IV, p. V. 13 Indeed, education remains Apple's most important long-term strategic market. More than 60 percent of the computers used in United States primary and secondary schools are Apple computers. Apple computers are in more than 75,000 primary and secondary schools in the United States along and on more than 3,000 college and university campuses around the world. 14 Schrage, "The Collaborative Organization," New York Times, November 11, 1990, Section 3, p. 13 15 Remarks of Alfred C. Sikes, Chairman, Federal Communications Commission, before the Washington Annenberg Conference on the 1992 World Administration Radio Conference, November 5, 1990, Washington, DC. 16 Id. 17 However, it is far from clear that digital cellular schemes now being described are intended to provide even the very limited utility of present analog cellular systems for conveying data. 18 See EUCO-COST group 231, WG3 proposal, among others, which describes 14 Mbps data links within a room. 19 GEN DOCKET NO. 81-413, adopted May 9, 1985. 20 902-928 MHz, 2400-2483.5 MHz, and 5725-5850 MHz. 21 Initially encompassed in 15.126. 22 GEN DOCKET NO. 89-234, adopted June 14, 1990. 23 See Apple's Comments on GEN DOCKET NO. 89-354. 24 An additional drawback is that, by the very nature of Part 15, Data-PCS would have only secondary status in the ISM bands and would have to cease operations if harmful interference were to result. 25 One informed observer recently commented (on operation under 15.247) that "external high power, narrow band licensed operations can disable communications from up to 17 miles away." Stuart J. Lipoff, ADL, at WINLAB Oct. 19, 1990. 26 The potential for unpredictable and uncontrollable interference from and to Part 15 devices is not theoretical. Even though only a handful of Part 15-based products have actually been introduced, operators of authorized services in the ISM band have already expressed concern. Teletrac, for example, has described "significant amounts of interference" from sources including Part 15 devices. Sensormatic has received temporary relief from interference from Part 15 devices for its operation in the 902-905 MHz range. 27 Reports that government radars in some locales occupy a portion of the 2400 MHz ISM band suggest further limits to the effectiveness of spread spectrum systems using these frequencies. 28 When in-room, line-of-sight connectivity is the dominant usage, higher frequencies such as the 18 or 60 GHz range and / or infrared may be preferred and practical. For longer distances of interconnection, such as across a campus, Data-PCS can bridge or gateway into other networks or media. 29 This range overlaps the UK's PCN spectrum of 1710-1785 and 1805-1880 MHz. 30 Many experts, however, believe that a much larger portion of the spectrum should be devoted to Data-PCS for wireless LANs. A subcommittee of the IEEE, for example, estimates the need as falling in the range of 70 MHz to 140 MHz. See IEEE 802.11 Reply Comments on GEN DOCKET 90-314. 31 For example, 10 Mbps (IEEE 802.3/Ethernet), 4/16 Mbps (802.5/Token Ring), etc. 32 Data-PCS will of course be appropriate for both high and relatively low data rates (e.g., Apple's LocalTalk(tm). Apple's petition is directed at providing usable radio spectrum for all wireless local area networks of today and the near future. 33 Because of the low RF power levels of Data-PCS, these 10 MHz segments could occupy the so-called "guard bands" requested in at least one PCS proposal. See PCN America (Millicom) Petition, dated November 7, 1989, FCC Docket RM-7175. 34 The lively dialogue between the Commission and industry on this subject when revising 15.247 was concluded by adoption of power limit expressions specific to different modulation schemes, including a power spectral density limit for direct sequence and an absolute power limit for frequency hopping spread spectrum. GEN DOCKET 89-354, June 14, 1990. 35 This provision will motivate directionality more than the one-for-one decrease required in 15.247, as revised July 9, 1990. There is no need for the Commission to require integrated antennas. In many applications, Data-PCS transmit antennas can be incorporated into the computing device, but use of a remote antenna (e.g., for remote peripherals) could enhance flexibility of usage and reduce the RF power required. 36 For example, transmitters which do not have associated receivers to determine channel usage, or which are capable of extended transmission periods without re-contention for the channel, must not be allowed. 37 The IEEE 802.3 standard calls for the equivalent of a maximum continuous time on channel of 1.235 milliseconds and a minimum time between transmissions of 9.6 microseconds. Apple believes that industry can develop and recommend similar specifications, as optimized to be applicable to the RF medium, to the Commission. 38 Descriptions of many CDMA spread spectrum schemes imply the need for power control of portable units through a feedback loop engaging a central node. Requiring CDMA may be tantamount to defining a central-node-based network architecture that could limit the effectiveness of Data-PCS. CDMA and other schemes, however, must not be prohibited by regulatory constraints. 39 Which amended Part 1 of its Rules implementing the National Environmental Policy Act of 1969 (NEPA). 40 Apple supports the IEEE 802.11 committee's efforts to emphasize intrinsic health and safety in addressing Medium Access Control (MAC) and Physical Layer (PHY) specifications for wireless LAN's. See Doc IEEE P802.11/90-20 - Proposal for PAR. 41 Wang Laboratories, Inc. 42 PR DOCKET No. 83-26. 43 Keynote speech to Educom '87.