4.1 What is a Network Computer?
First there were the mainframe computers, then came the personal computers (PC), now some say that the ``network computers'' (NCs) are the new entrants in the computer industry which will eventually edge out their predecessors. The most general definition of a network computer is a stripped-down computer connected via a network to a server which holds most of its software and data, and it can be stationary, nomadic (like a public telephone), or mobile (like a cellular phone). Proponents of network computers, headed by Oracle, Sun Microsystems, and IBM, contend that network computers will soon rival the PC industry and eventually edge it out. They argue that network computers are cheaper to own and to maintain, and offer more flexibility. A network computer by itself may be cheaper than a PC, but with the extra investment in the network connection and the server, the equipment costs on both sides weigh almost equally. The main saving, however, will probably come from a lowered administrative cost. A system administrator can manage more network computers than PCs since in the former case, only a few servers need to be maintained.A more compelling argument for investing in network computers is the additional flexibility that they are predicted to offer. As envisioned by their proponents, network computers are expected to replace mainframe terminals, PCs and telephones in the office, offer low-cost, easy-to-use PC-like devices at home, and provide a combination of mobile telephone and personal digital organizer on the street. That is, NCs are expected to offer users data access from anywhere.
In the offices, many mainframe terminals and PCs are used merely for data entry, and NCs may offer this capability and also connections to the Internet. Network Computer, Inc., a wholly owned subsidiary of Oracle, has already developed the system software and the server for NCs. Microsoft and Intel are rivaling with their own version, called the NetPC, which follows a similar computing paradigm, but with the additional requirement of containing Microsoft software and Intel chips. For the home users, WebTV offers set-top boxes which allow the consumers to surf the Internet via their TV sets and telephone lines. Such a set-top box costs $250, in comparison to more than $1000 for a cheap PC.
According to some analysts, the biggest market for NCs may be hand-held machines that combine into one the mobile telephone and the personal digital organizer. Nokia, a Finnish firm, is already selling such devices which consist of a simple computer with email, Web surfing, and common organizer tools. Currently, the speed of Internet connection is quite slow, but as the product becomes more sophisticated, this type of technology is expected to become more popular.
If network computers do become widespread, can technology keep up with the growing demand? The bottleneck would be the ability to provide fast Internet connection and data access from any location. Although these are mainly technological issues, the electronics industry does have to depend to a certain extent on the government to free up necessary resources before these technologies can be implemented. For example, to achieve the goal of ``access from anywhere'', it inevitably requires both wired and wireless access to be supported. While deregulation from the Telecommunications Act of 1996 has virtually allowed data transmission on any type of transmission lines (e.g. telephone, cable, and utility lines), wireless access requires frequency spectrum which is scarce and tightly managed by the government. Here we will examine what the regulators have done in recent years which may have potential impact on the wireless communications industry.
4.2 How might regulation affect its proliferation?
Besides low cost, one other main selling point of the network computer
is its capability to make Internet access readily available from any
location, that is, through both wired and wireless access. To achieve
this goal, there needs to be a network infrastructure which can
support a broad enough bandwidth. In the case of wired access, the
infrastructure is well on its way, especially after the deregulation
of the telecommunications industry. For example, with deregulation,
companies from different sectors such as the telephone and the cable
industry are developing new competing technologies to allow faster
modem access from homes and office buildings (
see the case study of
ADSL vs. cable modems). Even if traffic on the telephone and cable
lines becomes congested, it can always be alleviated by the addition
of more twisted pair wires or cables, although cost may be a major
impediment. In any case, impediment due to government regulation
concerning wired connection will be unlikely.
For wireless access, however, one does not have the luxury of a virtually unlimited resource for its infrastructure, since spectrum is scarce and regulated by the Federal Communications Commission (FCC). In response to the growing need of the wireless telecommunications market, the FCC has moved in the direction of market-driven and flexible spectrum management policies. To hasten the decision process of spectrum allocation and to encourage competition in wireless services which would benefit the public (and also to pour money into the Treasury), since the passage of the Omnibus Budget Reconciliation Act of 1993 the FCC has begun conducting auctions of the spectrum through several rounds of competitive bidding. This process is deemed to be better than the previous methods of lottery or comparative hearing because it is faster and it places spectrum into the hands of those who regard the spectrum valuable enough to pay a high price. In contrast, valuable time and energy is wasted when winners of spectrum lotteries resell their rights to other companies, or when comparative hearings becomes a long process especially when repeals are made to re-examine the case. In addition to the licensed spectrum, the FCC has also set aside spectrum for unlicensed low-powered local wireless devices (e.g. common household devices such as cordless telephones and garage door openers).
The FCC has recognized that while some amount of central management is necessary to avoid interference among users, spectrum users rather than a central planning body can best respond to changing market forces and needs. It has hence adopted a `flexible use' style of spectrum management, and intervenes only when such action would promote social benefits. In particular, this flexibility includes the freedom for the users to decide how they use the spectrum in terms of the types of services, technologies and equipments, the time line for implementation, and the freedom to decide how they want to partition or aggregate the authorized spectrum and the geographical area. The FCC is also open to proposals and comments from companies, in order to decide the best way to allocate or auction spectrum.
Even though the FCC has generally cooperated with market forces and demands, the spectrum pie is a much coveted treasure for which players from many industries are vying (not to mention also national security usages like the military or air traffic control). Can the FCC allocate enough spectrum to enable wireless digital communications to become ubiquitous? Below we will examine a few examples of recent cases of spectrum allocation that has already had or will have profound impact on wireless technology, and perhaps similar implications on the development of wireless network computers as well.
4.2.1 Allocating PCS spectrum
In several auctions commencing in December 1994, the FCC auctioned a
bandwidth of 120 MHz in the 2 GHz band for personal
communications services (Broadband PCS) in 51 Major Trading Areas
(MTAs) encompassing the entire United States and its territories.
Broadband PCS consists of ``a variety of mobile and/or portable radio
services - using such devices as small lightweight, multifunction
portable phones, portable fax machines, and advanced devices with
two-way data capabilities -- that are expected to compete with
existing cellular, paging and other land mobile services''.
[1] The goal of broadband PCS is to make the latest
telecommunications services available and affordable to the general
American public, which until recently were affordable only to business
users and the affluent. Among the top bidders were companies like
WirelessCo, AT&T Wireless PCS Inc, PCS PRIMECO, and Pacific Telesis
Mobile Services. In the past few years, the wireless
telecommunications industry has been burgeoning at a fast pace, due to
advances in technology and demands from users who have gotten used to
the convenience of Internet connection while sitting in a cafe, or
lazing on the beach. Indeed, with wireless modem one can use
applications like checking email and surfing the Web, although the
connection is quite slow. However, with the given bandwidth, wireless
PCS connection is not expected to accomodate broadband multimedia
services which transmit high quality graphics and video.
Due to the nature of the NII/SUPERNet devices, that is, its aptitude
in supporting short-range and high-speed LANs, this makes it ideal for
devices like the (wireless) network computers. For example, envision
that we have a computer terminal in every classroom of a school. The
maintainence of upgrading and updating each computer would be costly,
and hence the concept of having just a few central servers is very
appealing. Hence, if the network computer can gain grounds in
providing wireless LANs, it could potentially establish a stronghold.
Similar to the wireless cable system, the satellite system also has a
high up front cost and wide area coverage. There are quite a few
proposed satellite system, including Iridium Inc, backed heavily by
Motorola, who plans to launch 66 LEF satellites to link handheld
wireless phones over the entire globe. A more ambitious system is
proposed by Teledesic, backed by Bill Gates and Craig McCaw, where 840
LEO satellites will provide global coverage for two-way broadband
services, including voice, video, and multi-media. It will be years
before these systems become well established and widely used, and their
initial impact on wireless communications probably will
be minimal, since the high cost of such devices will be confined to a
niche market, such as connections for rural and remote areas, and a
small group of technophiles.
These two technologies are not yet well established. Nevertheless,
they do provide potential resources for network computers designed to
be nomadic or mobile.
In a recent development (April 1997), the FCC has set a timeline for
digital TV to appear by Christmas 1998, and to convert the nation's TV
broadcasting to completely digital by 2006, shortened from its
original 15 years timeline. The 18 month deadline is voluntary and
there is no penalty. As described in the Telecommmunications Act of
1996, the broadcasters are allowed flexible use on the additional
spectrum, and may offer ``subscription and pay-per-view programming,
wireless telephone service, paging, data transmission, or any other
service.'' In view of advancing wireless communication, this may be a
good news; in view of business strategies however, one can see that
they have an unfair advantage over the PCS companies, who have to
compete with each other for a share in the spectrum. At the time of
this article, there is no know plan for the broadcasters to venture
into the wireless computing business.
In contrast, companies in the PCS, wireless cable and satellite
industries must bid at high prices for their share of the spectrum
pie. This price tag may have proven to be too steep, as some PCS
companies are having difficulties paying the installments, and one
company has already declared bankruptcy. The FCC has temporarily
suspended the collection of these installments. If the spectrum were
to be re-auctioned, it is unlikely that it will bring in as much money
as the first auction. Moreover, this further delays the
development of technology.
In conclusion, we believe that government regulation will not pose as
an significant impediment to the advancement of wireless technologies,
in particular, the wireless network computer. The success or failure
of NCs will more likely be determined by other factors such as
technological issues, costs, and human factors. The cost savings that
NC endorsers have been hailing may be incentive enough for some
companies, especially those that have been frustrated by PCs or those
that would benefit greatly from the lowered administrative cost. On
the other hand, current NCs have very limited number of applications,
and switching to NCs would put most of the world's software (which are
made for PCs) out of reach. Besides the usual concerns of cost, speed
and quality of services, a more controversial topic would be the
debate on the suitability of the NC computing paradigm. PCs enabled
the transition from the centrally managed computing paradigm in
mainframes to user empowerment. Now NCs are trying to reverse this
direction, though promising more capabilities. For businesses where
central management of computer software and databases is desirable
(e.g. data entry and shared computing), NCs have great potentials in
gaining market. However, in sectors where users are accustomed to the
empowerment that PCs offer will be reluctant in switching to NCs, for
fear that their creativity and productivity be stymied and their
security be violated. Our prediction is that NCs will complement PCs,
and not replace them. More specifically, they will serve companies
where central computing is suitable, while the rest will remain happy
with PCs.
[1] FCC-WTB Auctions Fact Sheet"
4.2.2 Unlicensed spectrum for short-range high-speed
In response to proposals from the Wireless
Information Network Forum (WINForum) and Apple Computer, the FCC has
approved 300 MHz in the 5 GHz band for a new category of unlicensed
equipments in January 1997 (dubbed the NII/SUPERNet devices). These
devices are expected to provide short-range high-speed wireless
digital information transfer, and in particular is hoped to support
wireless local area networks (LANs) and connection to the National
Information Infrastructure (NII). Unlicensed devices have become
commodities at homes, like the cordless telephone, garage door opener,
security alarm sytem, and electronic toys. It has also become
prevalent in business, with examples like remote meter reading,
medical patient monitoring, and inventory control systems. This new
category of unlicensed wireless equipment is expected to support LANs
and access to the Internet, with major advantages over wired
access be its low-cost, mobility, versatility, and flexibility. In
particular, the proposals were originally envisioned to benefit
schools, libraries, and hospitals, but businesses will likely profit
as well.
4.2.3 Possible wireless alternatives?
There are some alternative wireless technologies which could
potentially be converted to two-way digital data communication,
namely, wireless cable and satellite. Wireless cable is originally
planned as an wireless alternative to cable television, and it
includes an upstream spectrum for interactive video services. A total
of 3 channels can be offered by Multipoint Distribution Service (MDS)
and Multichannel Multipoint Distribution Service (MMDS) licensees in
the 2.596-2.644 GHz band. In addition, the FCC allocated another 2
GHz in the 27.5-29.5 GHz band to the Local Multipoint Distribution
Service (LMDS). Both the MMDS and LMDS systems plan to use digital
technology to increase channel capacity and provide for limited
two-way interactive service. However, since these services were
originally designed for fixed locations, it is not know how suitable
it is for mobile devices. One disadvantage of wireless cable is that
it depends on line of sight, thus the signal quality can be degraded
by adverse weather and terrain (e.g. trees). Analysts expect that
upgrading wireless cable to a digital format should not be
difficult, but little is done about it so far. If wireless cable were
to play a role in providing mobile data access, it would be more likely
that it is used in combination with other mobile services such as PCS,
to provide a total wireless mobile system.
4.2.4 Obstacles Ahead?
Although the recent trend has shown policy makers to respond favorably to
the growing market of wireless communications, they still
shows inclination towards incumbents such as TV broadcasters. In
preparation for uphauling the nation's television broadcasting into
the digital phase, as recommended in the Telecommunications
Act of 1996, the FCC freely handed off a slice of the
spectrum to incumbent TV broadcasters. As a result, the FCC has
doubled the spectrum currently allocated to incumbent television
broadcasters to facilitate the transition to the digital technology
high-definition television (HDTV), which allows better quality images
and the simultaneous broadcasting of multiple channels. The intention
of this plan is to make broadcast TV more competitive to cable and
satellite services. However, this decision comes at a time when
broadcast TV is acting less and less for public interest, manifested
in the fact that many programs of sex and violence are aired despite public
outcries and fewer and fewer educational and community-originated
programs are being broadcasted. The controversy of this plan mainly
centers around the obligation of broadcasters to act according to
public interest (in terms of their programming content) and largely
irrelevant to the point of this paper. What is relevant is the fact
that policy makers so far seem more willing to allocate this scarce
resource to incumbents for the HDTV technology in which the public is
relatively uninterested, rather than looking ahead to allocate it to
potential new users such as Internet providers and wireless computer
networks which may benefit the public far greater in a novel way by
making Internet access and wireless computing more prevalent. In this
sense, the development of wireless data communications may be held back
by regulation since its main resource has been made more scarce,
forcing technologists to develop more advanced technologies (e.g.
better compression using smaller bandwidth) to cope with what they
have available. An explanation for this preferential treatment may be
that digital television technology is a relatively well-researched and
matured field, whereas wireless computing is still very much in the
developmental stage, and will be years before it becomes a commodity.
Perhaps it is due to the infancy stage of this
technology and the lack of urgency that has the policy makers
overlooking it when it comes to spectrum allocation.
4.3 What does this all mean?
Since even the wired network computers have yet to gain market, it is
premature to speculate on the future of wireless network computers.
With the recently allocated unlicensed spectrum, we are likely to
start seeing more local area networks of wireless connection. As for
surfing the Web or other broadband services on a mobile machine, it is
unlikely to achieve adequacy and popularity with the current allocated
spectrum (broadband PCS and wireless modem), unless the wireless cable
and satellite industry also show technological improvement in this
area. However, the FCC has shown consistent willingness to meet the
demands of burgeoning wireless services and would not likely pose an
impediment should it continue to grow in the future. Given the
current available spectrum and technology, it is more likely that
broadband wireless applications will be carried by several
complementary services, a possible scenario being the long distance
services carried by wirelines, which have smaller error rates,
and local connections served by wireless technologies.
Footnote
