Wireless Broadband 1
Running Head: WIRELESS BROADBAND ACCESS
Wireless Broadband Access
Jonathan B. Steele
July 10, 2003
University of Maryland University College
MSIT 660, Section 9040
Wireless Broadband 2
With the broadband revolution in full swing, the dialup modem will likely end up in the
technological graveyard with other treasures, like record players and eight track tapes. As
companies tried to leverage different technologies for market share, the wireless technologies
stumbled getting out of the starting gate while wired access technologies emerged as the
overwhelming front runners. Wireless technologies have since overcome some of the technical
and regulatory barriers that plagued first generation systems. A new breed of entrepreneur is
again excited about the prospects of bringing wireless broadband to the masses. Is the wireless
David now poised to take on the wired Goliath, or will wireless broadband be permanently
allocated the role of gap filler?
Wireless Broadband 3
This paper explores the technological and economic landscape of wireless broadband
access. The Federal Communications Commission (FCC) defines broadband as “… access to
Internet and Internet-related services at significantly higher speeds than traditional [dialup]
modems” (FCC, 2003). Other definitions indicate a threshold of 200 kilobits per seconds (Kbps)
in either direction as being the qualification for broadband.
The intended scope of this paper is residential wireless broadband access, but in order to
understand where wireless access has been and where it may being going, it will also be
necessary to briefly explore the wired technologies as well as non-residential wireless
applications that may eventually find their way into the home.
The paper will start with an assessment of some of the fundamental drivers behind the
popularity of broadband access in general. It will then touch on the wired access technologies
that are currently dominating the broadband market. The paper will identify some of the
difficulties encountered in trying to get the first generation of wireless access technologies to
market, focusing first on satellite-based systems, and then progressing into terrestrial-based
wireless systems. It will then show how wireless equipment manufacturers and service providers
have applied lessons learn to try and improve equipment and customer interaction. The paper
will then summarize the current state of residential wireless broadband and close with a forecast
on the strength of the rebound the industry might see.
The Broadband Revolution
Broadband’s popularity is stilling booming despite a somewhat depressed economic
environment. Dialup access presents a host of problems that consumers have suffered with since
the early days of the Internet. One of most significant issues is that it ties up the telephone line.
Many households get around this problem by leasing a second telephone line, which also comes
Wireless Broadband 4
in handy for households with teenagers. The other major issue with dialup connections is speed.
Today’s web content can be graphically intensive. While some providers are still offering low-
bandwidth versions of their web sites, this extra work is likely to be sacrificed as broadband’s
popularity continues to soar. Software providers are also putting increased reliance on the
Internet as a method to distribute patches, service packs, and in some cases, entire applications.
These downloads often require hours of online time and one disconnection means having to start
all over. Lesser issues associated with dialup can be categorized as nuisances resulting from the
temperamental nature of the connections.
On the broadband side, the primary advantages are the opposite of the dialup
disadvantages. With broadband, the access is always on and users can download files that used to
take 20 minutes, in a matter of seconds. The downside to all of this capability is money.
Broadband costs about four times as much as an economy dialup Internet Service Provider (ISP),
but add a second phone line and move up to one of the premium dialup ISPs, and the costs
become comparable to broadband. A secondary downside to broadband that is not mentioned a
lot, but important for frequent travelers, is that broadband customers are usually out in the cold
when they are away from home, whereas the bigger ISPs offer nationwide access to their
services at no extra cost.
On the business side of the house, residential communication offers an attractive target
for service providers. For families that want to stay plugged in with the state-of-the-art, it is not
unusual to spend up to and beyond two hundred dollars a month for local and long distance
telephone, cellular phone service, cable or satellite television, and broadband Internet access.
Logic mandates that service providers that are able to package several of these services together
with attractive incentives may be able to lure customers away from single-service providers. We
will see later in this paper that this strategy is already being employed and the trend is likely to
Wireless Broadband 5
The Big Gorillas
Cable-based high speed internet and telephone line-based Digital Subscriber Lines
(DSLs) have emerged as the Goliath in the residential broadband market. Juniper Research
forecasts indicate that by 2006, the combined cable/DSL market share will have grown to over
38%, from about 15% in 2001. This growth is reflected in a similar decline in the dialup market
share. The data also indicates that DSL, while still number two, is closing the gap on cable
(Telecommunications Industry Association, 2001).
Both the cable and local telephone companies have invested millions in upgrading their
distribution infrastructures, in some cases bringing fiber optics to the curb. On the surface, it
might appear that cable has a performance edge over DSL, but cable’s performance suffers as
additional neighbors sign up for service.
As DSL becomes more available and more reliable, the minor performance differences
between cable and DSL become inconsequential, and the distinguishing factor is likely to be
price. Cable companies theoretically have a head start for customers that are already subscribing
to cable television. Cable service price structures are typically tiered, providing a discount for
customers who subscribe to both television and Internet services. This approach can also be a
detriment, and is likely the main reason that DSL is closing the gap on cable. With DSL
matching or even beating the discounted cable Internet price, the choice becomes a no-brainer
for consumers that do not have or want cable television.
Wireless Access Technologies
Wireless broadband comes in two basic flavors: satellite-based systems and terrestrial-
based systems. Satellite-based systems can be further divided-based on the type of orbit the
satellites are in. Satellites in what is called geostationary (GEO) orbit appear to stay in the same
position in the sky. The position where this phenomenon occurs is at an orbit of over 22,000
miles directly above the earth’s equator. The main benefit of GEO orbits is that only a small
Wireless Broadband 6
number of satellites are required to provide broad coverage. Most of the earth can theoretically
be covered with only three satellites. The disadvantages of GEO satellites include higher launch
and deployment costs, additional complexity and power required to transmit such a long
distance, and the need for a southern exposure to be able to accommodate line of sight (LOS)
requirements. GEO satellites also introduce about a quarter of a second latency due to the travel
time between the ground and the satellites. This latency can complicate connection-oriented
protocols like TCP/IP, which is used extensively on the Internet. Low Earth Orbit (LEO)
satellites are closest to the earth, at orbital altitudes ranging from about 400 to 1,000 miles. LEO
satellites offer the most performance potential, but constant coverage requires numerous
satellites with sophisticated inter-satellite communications. Medium Earth Orbit (MEO) satellites
are located in between LEO and GEO, with orbital altitudes ranging from about 2,500 to 6,000
miles. MEO satellite networks require fewer satellites than LEO constellations, but are still much
more complex than GEO systems (Fowler, 1998).
Terrestrial-based wireless broadband must use radio transceivers located on the ground,
similar to how the cellular telephone networks have evolved. Terrestrial-based wireless
broadband options can be categorized based on the radio frequency that the service is using. The
two broad categories of frequency spectrum are licensed and unlicensed. Examples of devices
that use licensed spectrum are cell phones, police radios, and analog radio and television.
Devices like cordless phones, walkie-talkies and wireless LANs use unlicensed frequencies.
The licensed versions of terrestrial-based wireless broadband can be further divided into
two categories called Local Multipoint Distribution System (LMDS) and Multichannel
Multipoint Distribution System (MMDS). LMDS uses the radio spectrum between 24 and 31.3
gigahertz (GHz) whereas MMDS uses the spectrum between 2.15 and 2.68 GHz (Ortiz, 2000).
Generally speaking, lower frequencies have better propagation properties than higher ones. Their
radio waves cover broader areas, can penetrate obstacles better, and are more immune to nature’s
Wireless Broadband 7
fury, like heavy rain. As an extreme testament to this theory, the Navy uses ultra-low frequencies
to communicate with submarines at sea. As a result, LMDS has a small service footprint, with
cells only one and a quarter to two and a half miles wide. The technology required to support
LMDS is more sophisticated and the ground equipment is more expensive. The advantage of
LMDS is that the government has allocated the spectrum such that there is plenty of bandwidth
that can be applied to support higher data rates than MMDS. LMDS can theoretically support
data rates of up to 1.54 gigabits per second (Gbps) downstream, but services are typically offered
at 45 megabits per second (Mbps), with higher rates on the horizon (Elliott, 2001). This
combination of properties makes LMDS better suited to business applications than residential. A
typical LMDS deployment might be for a small business campus, where multiple buildings can
be serviced by a LMDS connection.
The MMDS frequency spectrum was originally intended for subscription-based
television, but the technology couldn’t compete with the cable and satellite television providers.
Recognizing that the original plan for the MMDS spectrum didn’t pan out, the FCC amended
regulations in 1998 to allow for two-way use of the MMDS frequency spectrum. This opened up
the door to other applications beyond broadcast television. MMDS cells can be up to 35 miles
wide, so larger geographical areas can be covered with less equipment. The ground equipment is
less complex than LMDS, which helps reduce manufacturing costs. The drawback to MMDS is
that the FCC allocated the spectrum such that providers have smaller frequency ranges to work
with. A single MMDS channel can theoretically support data rates up to 27 Mbps, but this
bandwidth must be divided among recipients, resulting in real data rates in the range of 300 Kbps
to 3 Mbps. These combined characteristics make MMDS the logical choice for residential
applications (Ortiz, 2000).
The two most popular bands of unlicensed spectrum are the Industrial, Scientific and
Medical (ISM) band and the Unlicensed National Information Infrastructure (U-NII) band. ISM
Wireless Broadband 8
includes bandwidth in the 900 megahertz (MHz), 2.4 GHz and 5.8 GHz ranges. U-NII provides
additional bandwidth in the 5 to 6 GHz range (Mead, 2001). While there are some limitations
associated with using unlicensed spectrum, to be explored a little later, it does offer an attractive
alternative for potential wireless broadband applications.
A Rough Start
The same Juniper Research study quoted earlier shows hardly any market share growth
for the combined wireless broadband technologies between 2001 and 2006. While the wireless
technologies lured investors anxious to capitalize on the high-tech upswing, providers soon saw
that all was not paradise in Eden.
While GEO satellites had been providing communications services for years, the late
nineties brought a new generation of LEO and MEO satellite constellations that promised to
eliminate dreaded GEO latency, and revolutionize voice and data services for everyone.
Unfortunately, these systems were very expensive, provided mediocre performance and could
not attract the customers necessary to sustain operations. The bankruptcy and near demise of
Iridium capped a freefall of these new generation systems. While still operational today because
of a last minute bailout with government intervention, Iridium’s technology could not compete
with ground-based advances (Steele, 2000). While Iridium and several of the other early systems
focused on voice communications, their failure had significant implications on other
constellations that had been on the drawing board and would have been capable of providing
After the string of new fangled system bankruptcies eased a bit, GEO satellites once
again emerged as the vehicle of choice for the broadband market. Like its cable cousin, satellite-
based broadband is linked inexorably to satellite-based television. Hughes’ DirecPC system
leveraged its DirecTV broadcast television technology to provide half a broadband capability for
downloading large files. Uploads still had to be handled by a terrestrial-based ISP. While the
Wireless Broadband 9
system worked, it did not resolve all of the irritating aspects of having to use a dialup connection.
In addition, the ground equipment required to communicate with a GEO satellite was very
expensive and hard to set up. The same installation teams that installed satellite television
antennas were expected to install the broadband antennas with little or no additional training.
The installations sometimes doubled or even tripled the assigned times and often required return
trips to stabilize a system (Ferguson, 2001). To add insult to injury, the customer base was
further constrained because of the need to have a clear line of site to the south since GEO
satellites are all positioned over the equator. So much for wooded lots or the north sides of
The state of the satellite television industry combined with a slow economy set General
Motors in motion looking for a buyer for its Hughes Electronics holding. Hughes appeared to be
poised for a merger with another satellite television powerhouse, EchoStar, but the FCC blocked
the merger citing the impact of reduced competition in satellite television’s monopoly of rural
America, where cable was not yet available. The companies argued in vain that the efficiencies
that would be gained through the merger were necessary to compete with cable television
providers (Wolverton, 2002). This blow to the satellite television industry also slowed
momentum in the roll out of satellite broadband capabilities.
On the terrestrial-based side of the house, major telecommunications companies like
Sprint, WorldCom and AT&T were bolstered by the success of their cellular phone endeavors
and envisioned a wireless revolution. Their goal was for MMDS to be a competitive “third pipe”,
along with cable and DSL, in providing broadband Internet access. They also saw opportunity in
areas where cable and DSL were not available. When the FCC auctioned off MMDS spectrum in
the mid-nineties, a frenzied bidding war resulted (Suydam, 2002).
Armed with their expensive bandwidth, these telecommunications giants proceeded to
rollout MMDS service in markets around the country. The only problem was, they weren’t
Wireless Broadband 10
making money; not by a long shot. Two terms that are prevalent in many of the articles
concerning the collapse of the MMDS industry are “truck rolls” and “self installation.”
Trucks rolls are individual service calls to residential customers. An ideal scenario is to
have customers install the equipment themselves, thereby avoiding any truck rolls. This was not
an option for the first generation MMDS equipment because they were LOS-based systems that
required professional installation to align and optimize the signal. The next best scenario would
be one truck roll to install and test the MMDS equipment and leave the customer satisfied. Sprint
was averaging three truck rolls to get a single customer up and running, at an astounding
combined labor cost of over $1000. The equipment was temperamental and even after getting a
system working, it was not unusual to have to roll a truck again because the wind or some other
culprit bumped an antenna (Suydam, 2002).
MMDS’s problems were compounded by the fact that customers weren’t happy with the
service they were getting. Download speeds were just a fraction of what had been promised and
sometimes approached that of dialup connections, perhaps due to over-subscription. System
tinkering impacted reliability leading to threatened lawsuits. Speculation was that Sprint’s
attention was focused more on next generation MMDS technology and not fixing the seemingly
insurmountable problems inherent in the first generation deployments (Greenfield, 2001).
In the unlicensed realm, an innovative company called Ricochet deployed a 900 MHz-
based system that used antennas mounted on street lights to provide near-broadband speed
Internet access. At its peak, Ricochet served 13 cities with 128 Kbps service. When Ricochet
stopped operations in August 2001, their 51,000 customers amounted to only about one fifth of
what would have been required to keep the company solvent (Cherry, 2002).
Help On The Way
The major detractor facing the satellite broadband industry was the continued reliance on
a dialup connection for uploads. Two areas of research helped overcome this reliance. The first
Wireless Broadband 11
area of research focused on the TCP/IP protocol itself. TCP/IP is the prominent protocol used on
the Internet. It was designed for terrestrial networks with relatively little latency, but lots of
congestion. Satellite link characteristics are just the opposite, with little congestion, but lots of
latency, particularly with GEO satellites. TCP/IP incorporates flow control techniques which are
not suited for the long delays. Special protocol gateways were designed to alter TCP/IP
operations to better fit the satellite link model. Modifications included expanding flow control
windows and sending multiple copies of packets to battle interference. Standard TCP/IP packets
would arrive at a protocol gateway which would use the satellite friendly protocol to send the
packets to the satellite. The satellite would then relay the packets to the ground equipment
installed at customer locations where another gateway would convert the packets back to
standard TCP/IP (Goldman, 2000).
The second area of research that helped break the one-way barrier was advances in the
ground equipment that needed to be installed at the customer location. An Israeli company called
Gilat Satellite Networks was successful in integrating a transmitting capability into its Very
Small Aperture Terminal (VSAT). Gilat was able to configure this two-way VSAT in a package
that was considered reasonable for small business or residential applications, with a dish antenna
diameter of about two to four feet (Gilat, 2003).
Returning to the lessons learned from the first generation terrestrial wireless systems,
new technology is emerging to solve coverage and performance issues associated with LOS
constraints. The new technology, dubbed non-line-of-sight (NLOS) incorporates smart antennas,
advance modulation techniques and mesh architectures that use customer equipment as relay
nodes, similar to how routers relay packets on the Internet.
The smart antennas use small arrays that can form individual beams directed towards
customers whereas first generation technology broadcasts signals over a much bigger area. This
technology focuses more power to the individual subscribers so that the positioning of subscriber
Wireless Broadband 12
equipment becomes less critical. The advanced modulation techniques include orthogonal
frequency division multiplexing (OFDM), which greatly improves spectral efficiency over first
generation techniques (Schrick & Riezenman, 2002).
Another interesting pursuit in next generation wireless systems is to take what used to be
a detriment to wireless systems, multipath, and turn it into an asset. Multipath is a type of radio
frequency distortion that results from signals arriving at a receiver at different times. Direct LOS
signals arrive first, followed by other versions of the signal that take alternate routes to get to the
receiver, like bouncing off of buildings. New digital signal processing technologies have been
developed to recognize the multipath signatures by analyzing the signals coming from individual
subscribers. These signatures are then used to compute the proper way to transmit the download
signals in such a way that alternate versions all arrive at the destination at the same time (Schrick
& Riezenman, 2002).
Several companies are researching mesh architectures. The basic concept is to equip
residential nodes with wireless routers that accept signals intended for a particular residence and
then retransmit the other signals to neighboring residences. Contrasting this approach to the
Ricochet system described earlier, it also solves the right-of-way and leasing arrangement
headaches associated with using municipal assets. While trial installations have been successful
in California, detractors of this technology point out that router density is critical, particularly if
the system is using ISM frequencies that limit the transmitter power of the wireless routers. They
also point out the potential latency implications associated with adding hops to the routing path
(Schrick & Riezenman, 2002).
This combined new technology goes a long way towards solving most of the
inadequacies of the first generation systems. The NLOS design opens up the possibility for broad
coverage within a service area whereas previously only partial coverage was possible because of
LOS restrictions. It also promises to reduce installation difficulties since technicians won’t have
Wireless Broadband 13
to tune the directional antennas. It even puts customer installation within the realm of possibility.
That being said, it’s difficult to ascertain how much of this new technology is real and how much
is vendor hype. Even if the technology is real, is it affordable? Service providers are not likely to
be lured by sexy technology without a viable business case, particularly after hemorrhaging
money in their first generation ventures.
The Current Climate
In the satellite arena, Gilat, armed with their new VSAT technology, actively pursued
partners that could help promote large scale applications. Satellite television giant EchoStar
invested in Gilat in 2000 and the combined team became the first to offer two-way satellite
Internet access, thereby allowing subscribers to deep six their dialup modems and all of the
headaches that go along with them (Gilat, 2000). The partnership, which also included
Microsoft, later emerged as StarBand Communications. This strategic alignment of
complimenting capabilities was able to effectively mount a challenge against Hughes’
dominance in the satellite broadband market.
Not to be outdone, Hughes was subsequently able to successfully integrate its own VSAT
technology into its DirecPC systems to enable two-way communications. The new service was
rebadged as DirecWay to distinguish it from the first generation one-way system.
At the end of 2001, DirecWay and StarBand had about 150,000 customers, but were
hoping to tap into their cable television customer base of over 10,000,000 (Marek, 2002). While
DirecWay and StarBand both use Ku-band frequencies, other startup companies are planning to
storm the market with Ka-band GEO satellites, which will handle up to four times the capacity of
their Ku siblings (Ferguson, 2001). WildBlue has emerged as a leader in the Ka-band field and
plans to initiate operations in 2004 (WildBlue, 2003). Hughes also has Ka-band plans with yet
another system moniker called SpaceWay (Hughes Network Systems, 2003).
Wireless Broadband 14
In addition to these established and planned GEO systems, other types of systems, like
Teledesic, which is a MEO-based satellite network, are still on the drawing board. Teledesic is
backed by deep pockets including Microsoft’s Bill Gates and wireless communications pioneer
Craig McCaw. Their web site indicates their service will start in 2005 (Teledesic, 2003).
So life after the failed Hughes-EchoStar merger continues. Judging by the amount of anti-
satellite advertising being sponsored by the cable television industry, it appears that at least the
television side of the satellite industry is making inroads into cable’s market share. The
broadband side of the satellite industry is still in its infancy. Despite the apparent optimism of the
satellite service providers, recent history suggests that they will have an uphill battle. While the
two-way capabilities solved one problem for the industry, it exacerbated another in that
residential equipment is now even more complicated because a satellite uplink transmission
capability is required in addition to the reception capability. This complexity translates into
dollars, and these dollars still must be passed on to the consumer. The service providers are
trying to lessen the blow of the initial $400-$600 investment by allowing customers to spread the
payments out over time, or offering reduced costs if the customers sign a long term contract
On the terrestrial side of the house, MMDS appears to have disappeared from the North
American wireless broadband market landscape. Sprint’s and WorldCom’s web sites do not even
mention fixed wireless service. While second generation MMDS technology is emerging to
resolve many of the issues encountered during the initial rollout, the technology does not
necessarily solve the MMDS business case. The big players are once burned, twice shy and have
adopted a “show me” attitude. Marginal improvement will not do the trick. Vendors are going to
have to roll out sure-thing solutions on a silver platter if MMDS is to make a significant
comeback (Marek, 2001).
Wireless Broadband 15
In an effort to recoup some of their investment, MMDS license holders lobbied the FCC
to consider alternate uses of the MMDS spectrum. In September, 2001, the FCC issued an order
setting the stage for the use of MMDS spectrum for “new advanced wireless services”, including
third generation (3G) mobile services (FCC, 2001).
There are still some smaller MMDS services sprinkled throughout the U.S.; probably
where the geographical terrain is particularly well suited for wireless transmissions, but the goal
of being a “third pipe” in the residential broadband market does not appear to be viable. MMDS
has faired a little better in overseas markets where the wired infrastructure is decades behind that
of the United States.
A trendy wireless service that emerged after the MMDS debacle has been coined
Wireless Fidelity (WiFi). WiFi uses unlicensed frequency bands and leverages wireless LAN
technologies to provide service to small cells called hot spots. In fact, a residential wireless LAN
can indeed be considered its own hot spot (Schrick & Riezenman, 2002). By using daisy-chained
access points, the same wireless LAN could not only serve the computers within a given
residence, but also the neighbor’s computers. If more neighbors installed access points, the entire
neighborhood could use the service. This is the concept behind WiFi, except the target customers
are generally commercial rather than residential. WiFi is currently being deployed in airports,
hotels, coffee shops and even throughout portions of downtown Manhattan (Shim, 2003). WiFi
service is provided by Wireless Internet Service Providers (WISPs). An example deployment
might be in a four story hotel. The hotel contracts with a WISP to provide connectivity to the
Internet through a DSL line or other terrestrial means. The WISP then installs the wireless LAN
infrastructure necessary to reach all of the guest rooms. The WISP charges the hotel a monthly
fee for the ISP service and for maintaining the wireless LAN equipment. The hotel then charges
each guest who wants to connect to its wireless LAN a daily surcharge. The same concept works
well for airports, where business travelers are anxious to be productive during layovers and flight
Wireless Broadband 16
delays. The business model makes sense if enough customers use the service, and with the
booming popularity of wireless LANs at home, it may quickly become viable. Other businesses
are providing the service for free as a way to bring customers into their establishments. The
investment in a DSL service and some wireless LAN equipment hardly exceeds petty cash levels
(Smetannikov, 2002). One can also envision commercial building owners going to WiFi versus
the cost of wiring the entire building for standard Cat-5 Ethernet access. Ah, but the focus of this
paper is supposed to be residential access. The WiFi model could also fit densely populated
residential dwellings such as high-rise apartment buildings. It should also be noted that the
aforementioned mesh network is very closely related to the WiFi initiative, and very well might
end up being the residential incarnation of WiFi.
Further evidence of the evolving “hot spot” approach can be gained by returning to the
Ricochet case study discussed earlier. Aerie Networks bought Ricochet and all its assets out of
bankruptcy for fire sale prices in 2001. Like Iridium, they have been temporarily reincarnated
without the burden of having to pay back development and deployment costs. Aerie is now
pursuing more of a hot spot model than broad coverage (Cherry, 2002). Their web site advertises
service in Denver and San Diego (Ricochet, 2003). They would appear to face an uphill battle;
particularly with WiFi emerging as potential competition for densely populated markets.
With most of the interest in terrestrial wireless broadband service currently focused in the
unlicensed spectrum, it is worthwhile to take a little closer look at some of its constraints. In
addition to well known devices that operate in unlicensed bands, a host of lesser known devices
also take advantage of the free airwaves; like microwave ovens, baby monitors, wireless audio
speaker system, and security alarm systems. As wireless LANs continue to gain popularity, the
scenario of saturation of unlicensed bands can’t be ignored. The very thing that makes the
unlicensed bands attractive to wireless broadband service providers, also represents a very real
business risk. Investors should be leery of putting all of their eggs in one basket without the
Wireless Broadband 17
safety net of licensed bandwidth behind them. If over-saturation ultimately impacts the
performance of their wireless broadband services, they will have no recourse and will ultimately
fail. In order to avoid this doomsday saturation scenario, the FCC imposes strict guidelines in the
use of unlicensed spectrum. Among the more substantial of these guidelines are power
limitations and the use of spread spectrum technology. While these measures definitely help
mitigate interference, they may not be able to withstand the onslaught of exponential growth as
more devices and users proliferate. The power limitations also constrain the coverage areas
thereby making it difficult to leverage unlicensed technology to the same extent as MMDS
There has been some talk about the much ballyhooed 3G wireless technology expanding
out of its mobile device niche and into the residential market (Palenchar, 2000). Proponents
suggest that there is no technical reason why a 3G infrastructure can’t communicate with home
gateways as well as mobile devices. Opponents point out that 3G uses shared infrastructure that
must be managed to avoid over-subscription (Drucker, 2001). The same can be said for
managing today’s mobile voice and 2G environment. That is why service plans only provide a
paltry amount of “anytime” minutes. Pricing structures for emerging 3G services in Europe and
Japan often apply a per-packet or per-minute surcharge in addition to a monthly fee. It’s one
thing for someone to download an occasional MP3 file to their mobile device, but it’s an entirely
different animal to expect 3G to handle sustained residential traffic, like downloading a 30
megabyte (MB) operating system service pack. That being said, the FCC’s decision to allow
alternate uses of the MMDS spectrum effectively blurs the line between 3G and fixed wireless
applications, and provides license holders the best of both worlds as they wait to see which way
the technological winds blows.
Wireless Broadband 18
The Future Climate
It is reasonable to forecast that wireless broadband will always be around to fill coverage
gaps and support unique niche applications. In order to assess the possibility of wireless taking a
more prominent position in the residential broadband market, it is necessary to shift the
discussion back to the wired competition. In researching the trials, tribulations and renewed
hopes of the wireless broadband industry, four primary contrasts between wired and wireless
The first and perhaps most important contrast is that of the transmission medium itself.
Wireless technologies use radio frequency spectrum that is regulated. Licensing auctions have
escalated bidding wars on spectrum which makes wired infrastructure costs pale in comparison.
While the use of unlicensed frequencies has been successful recently, almost all articles warn of
saturation of the public frequencies if exponential growth continues. The medium is further
constrained in that only a proportionally small part of the spectrum is appropriate for wireless
broadband applications. While the wired technologies have significant infrastructure costs to deal
with, the wired medium allows for highly scalable architectures that are not bound by discrete
The second contrast is in the complexity behind the technology. Wireless transmissions
must deal with sophisticated antennas, amplifiers, frequency converters and other radio
frequency equipment in addition to the modem technology that both wireless and wired
transmissions share. This added complexity translates directly into added cost, so wireless
equipment will always be more expensive for the consumer, regardless of how creatively the
service providers try to hide these costs.
The third contrast is in how the technologies can leverage their infrastructure for dual-use
applications. For the wired technologies, the dual-use paradigm is undeniable. The same cable
can be used for both television and broadband, and the same wire can be used for telephone
Wireless Broadband 19
service and DSL internet. The dual-use fuels revenue potential and allows service providers to
invest profits from one technology to build out the other. While there is some synergy between
satellite television and satellite broadband, incorporating a satellite-based upload capability was
clearly an independent effort and a significant departure from leveraging standard satellite
broadcast television technology.
The final contrast is that of simple momentum. As the wired technologies take bigger and
bigger chunks of the dialup market share, more manufacturers are building cable and DSL
modems and their prices continue to drop. As local phone and cable companies see successful
dividends from neighboring companies’ infrastructure build-outs, they want to get in on the
action and are more willing to make infrastructure investments themselves. Meanwhile, the
wireless technologies seem to bask in only small victories and have seemingly acquiesced to
their role of playing second fiddle. An analogy of this kind of momentum might be that of
Microsoft’s domination of the personal computer operating system market. While many people
sing the praises of other operating systems like Macintosh or Linux, they have not been able to
break into the main stream because Microsoft has successfully established Windows as the de
facto operating system for personal computers.
The residential consumer demands two things when subscribing to a broadband internet
service: performance and value. Wireless broadband internet access technologies have failed to
keep pace with the wired technologies in either of these categories. While wireless technology
will continue to mature, the tremendous momentum of the wired technologies, combined with
the constrained characteristics that are inherent in wireless technologies represent hurdles that
will prevent wireless from joining its wired big brothers as a major player in the residential
Wireless Broadband 20
Cherry, S. (2002, March). What went wrong at Ricochet? IEEE Spectrum, 39 (3), 60-61.
Retrieved June 9, 2003, from:
Drucker, E. (2001, April 16). The need for speed. New fixed-wireless services move to compete
with wired markets. Retrieved June 10, 2003, from:
Ellliott, C. (2001, January). Building the wireless Internet. IEEE Spectrum, 38 (1), 14-16.
Retrieved June 9, 2003, from: IEEE Xplore database accessed through the National
Aeronautics and Space Administration (NASA) Goddard Space Flight Center (GSFC)
Federal Communications Commission (2001, September 24). First report and order and
memorandum opinion and order. Retrieved June 21, 2003, from:
Federal Communications Commission (2003). Broadband high speed Internet access. Retrieved
May 29, 2003, from: http://www.fcc.gov/cgb/broadband.html
Ferguson, J. (2001, October). Achieving orbit. Boardwatch, 15 (13), 48-51. Retrieved June 10,
2003, from: ProQuest Computing database accessed through UMUC Information and
Fowler, D. (1998, December). Satellites: the new bandwidth-busters? NetWorker, 2 (5), 5-10.
Retrieved May 29, 2003, from: ACM database accessed through UMUC Information and
Gilat Satellite Networks Ltd. (2000, April 10). Echostar investment press release. Retrieved June
23, 2003, from: http://www.gilat.com/PressRoom_PressRelease.asp?sbj=303
Wireless Broadband 21
Gilat Satellite Networks Ltd. (2003). Satellite basics. Retrieved June 23, 2003, from:
Goldman, A. (2000, July 19). Satellites reach the rest of the world. Retrieved June 23, 2003,
Greenfield, D. (2001, April 5). Special report: fixed wireless services: wonder or woe? Retrieved
June 8, 2003, from: http://www.networkmagazine.com/article/NMG20010321S0011
Hughes Network Systems (2003). Welcome to Spaceway. Retrieved June 23, 2003, from:
Marek, S. (2001, October 29). Fixed wireless setbacks are sending U.S. vendors scurrying for
better solutions and more promising markets. Retrieved June 13, 2003, from:
Marek, S. (2002, February 4). Satellite survivors spreading broadband's benefits. Retrieved June
13, 2003, from:
Mead, M. (2001, October). Finding fortune in unlicensed spectrum. Retrieved June 21, 2003,
Ortiz, S. (2000, July). Broadband fixed wireless travels the last mile. IEEE Computer, 33 (7), 18-
21. Retrieved May 29, 2003, from: IEEE Computer Society Digital Library database
accessed through UMUC Information and Library Services.
Palenchar, J. (2000, September 25). 3G, home gateways are perfect together. Retrieved June 10,
Wireless Broadband 22
Ricochet (2003). Ricochet network coverage area. Retrieved June 14, 2003, from:
Schrick, B. & Riezenman, M. (2002, June). Wireless broadband in a box. IEEE Spectrum, 39 (6),
38-43, Retrieved June 9, 2003, from:
Shim, R. (2003, May 14). Wi-Fi as weapon—telecoms vs. cable. Retrieved on June 9, 2003,
Smetannikov, M. (2002, April 14). Wanted: a few good Wi-Fi POPs. Retrieved on June 5, 2003,
Steele, J. (2000, December 4). The rise and fall of Iridium. Retrieved on June 13, 2003, from:
Suydam, M. (2002, March 1). Fixed: broken? Retrieved June 9, 2003, from: http://www.e-
Telecommunications Industry Association (2001, November 12). The broadband household.
Retrieved June 14, 2003, from:
Teledesic (2003). About Teledesic. Retrieved June 8, 2003, from:
Wildblue. (2003). Answers to frequently asked questions. Retrieved June 2, 2003, from:
Wolverton, T. (2002, October 10). FCC votes no on satellite TV merger. Retrieved June 10,
2003, from: http://news.com.com/2102-1033-961706.html