Electronic Design
Broadband Communications Cut The Cord

Broadband Communications Cut The Cord

Broadband is going mobile. That means wireless, and it’s happening as more consumers access the Internet via laptops, smart phones ,and tablets. And that’s not all. Video has become the number one go-to app for broadband. Consumers want more video on their mobile devices as broadband service providers scramble for bandwidth and better technology to make it happen.

A broadband connection means fast Internet service typically from one of the established telecom carriers or cable companies. In the wired world, phone companies provide DSL service and cable TV providers offer DOCSIS-based service at speeds from 1.5 Mbits/s up to roughly 50 Mbits/s. Some companies also provide fiber to the home (FTTH), a fiber-optical service providing data rates to 100 Mbits/s.

In general, broadband is any speed greater than 1.5 Mbits/s, but that definition keeps changing as systems and products gradually get faster. All of these broadband services are fixed, meaning they aren't mobile. Furthermore, these services do not reach millions of potential subscribers in small rural towns and remote areas where it has been uneconomical to provide them to a relatively small audience. These millions are part of the so called Digital Divide, which lacks Internet service other than via the painfully slow dialup modem.

Then there’s wireless broadband—fast Internet connectivity that rivals wired services using radio technology. This is a growing category mainly because of its mobile nature and because new fixed services are becoming available. Wireless promises not only to fill in the gaps in broadband coverage but also to make ever more services available to smart phones, tablets, and other mobile devices (see “What’s Driving Wireless Broadband”).

Defining Wireless Broadband

Any wireless broadband service can be either mobile or fixed. Fixed means the user is in a single location like at home and is not moving. Mobile implies a moving target client such as a person with a smart phone, tablet, or laptop that is in motion. A modified definition of mobile is nomadic, which is a term describing a portable device that is fixed, at least for a time, such as a laptop at a coffee shop.

Broadband wireless service can serve all of these targets under the right conditions. Some technologies were optimized for mobile, while others are best at fixed or nomadic service. Mobile operation requires a technology that can adapt to the motion and provide handoffs between basestations as required.

Mobile data rates have continuously increased over the years as new technologies have emerged and as new products and services have come online. Early broadband wireless was hard pressed to reach the 1- to 2-Mbit/s range, but rates in the mid- to high-single-digit Mbits/s are common today. As new Long-Term Evolution (LTE) and other services come online, rates to 50 Mbits/s and above will be commonplace.

More than a few technologies are vying for a piece of the wireless broadband business. These include the familiar 3G and 4G cellular technologies, WiMAX, Wi-Fi, and several other proprietary systems.

3G/4G and Our LTE Future

The largest sector of wireless broadband comes from your local cellular provider in the form of 3G or 4G data services. 3G is almost fully implemented as carriers turn on the maxed out 3G HSPA+ systems. Concurrently, we are seeing the rollout of 4G, or what carriers are calling 4G. The major movement now is the expansion of LTE networks across the country.

Verizon is certainly the leader here with coverage in most major cities. Other smaller carriers like MetroPCS have also turned on LTE in selected areas. AT&T is still milking the HSPA 3G investment but has also implemented a few LTE stations with many more to come next year.

T-Mobile is staying with HSPA for the time being but no doubt has longer-term LTE plans. However, its fate is still up in the air as the Federal Communications Commission (FCC) and Justice Department are still debating the potential T-Mobile purchase by AT&T. It may be a while yet before we get the government’s final decision. An AT&T/T-Mobile combination would seem to be a positive development for wireless broadband as it would certainly extend the current coverage of AT&T cellular to better than 90% of the U.S.

LTE is an amazing wireless technology. It uses orthogonal frequency-division multiplexing (OFDM) and has just about the highest spectral efficiency of any wireless method. It produces more bits per hertz of bandwidth thanks to the wide spectrum it occupies and the higher-level modulation methods like quadrature amplitude modulation (QAM) it uses.

Using 64-state QAM (64QAM) in a 20-MHz channel, the maximum data rate can be 100 Mbits/s—and that’s with a single transmit stream. Using multiple-input multiple-output (MIMO) with multiple spectral streams, even higher rates can be achieved (see “Repealing Shannon’s Law To Boost Speed,” p. xx).

Using 2x2 MIMO, the maximum data rate can reach a peak of 172.8 Mbits/s under ideal conditions. Furthermore, OFDM with MIMO greatly minimizes multipath interference and in fact uses it to its advantage. This makes an LTE link very robust for mobile operations.

As for access method, standard LTE uses frequency-division duplex (FDD) with matching up/down channels. The downlink uses orthogonal frequency-division multiple access (OFDMA), and the uplink uses single-carrier frequency-division multiple access (SC-FDMA). Spectrum-saving time division duplex (TDD) versions of LTE are also available, although not used in the U.S. or Europe.

Even greater performance is on tap for the future with LTE-Advanced, the next level of the International Telecommunications Union (ITU) standard on the roadmap. Using higher levels of MIMO (4x4 and 8x8) and aggregating multiple 20-MHz channels, data rates to 1 Gbit/s are possible.

LTE is the technology to beat, and that’s why virtually all cellular carriers (with a few exceptions) throughout the world have chosen LTE as their 4G strategy. Look for LTE to be your wireless broadband link going forward (see “Wireless Companies Follow The Roadmap Past 4G And On To 5G,” June 23, p. 28).

As a further indication that LTE is our future, ABI Reserch analyst Phil Solis indicates that by the end of 2011, more than 200 million people in the U.S. will have access to an LTE network, and that’s expected to rise to more than 305 million by the end of 2016. An LTE phone, tablet, or laptop is probably in your future.

WiMAX Hangs In

WiMAX, which stands for Wireless Interoperability for Microwave Access, is a wireless metropolitan-area network (MAN) technology. As a 3G technology, it was available long before some other 3G services and LTE. But it has yet to hit that magic critical mass needed to make it the wireless broadband champ. It’s widely used outside the U.S, though.

WiMAX uses OFDM and time division multiple access (TDMA). It’s encapsulated in IEEE standard 802.16. The most widely used version, 802.16e, supports both fixed and mobile operation. Fixed basestations can support a range to 30 miles, while in mobile operation the typical range is in the 3- to 10-mile radius. The actual range depends on line-of-sight (LOS) or non-line-of-sight (NLOS) operation, frequency of operation, and tower height.

Frequencies of operation depend on the country and the carrier’s spectrum holdings. In the U.S., 2.5 GHz is the most common for licensed spectrum, but 2.3 GHz is also available for some. The U.S. also has 3.65-GHz spectrum. Worldwide, the 3.5-GHz band is the most common. Unlicensed operation typically uses the 5- to 5.8-GHz bands. WiMAX is not currently defined for the cellular bands like 700 MHz, 1.7, and 2.1 GHz, but it is an option.

The maximum rate data for a single 20-MHz channel using 64QAM is 75 Mbits/s, which is usually divided up into smaller, slower channels. The modulation is adaptable to range and other conditions and could drop back to 16QAM or quadrature phase-shift keying (QPSK) with too much noise, excessive range, or an NLOS problem. In any case, in practical systems the rate is usually 1 Mbit/s minimum up to 6 Mbits/s or more under favorable conditions.

Most WiMAX systems use 10-MHz channels and operate in the 2.5- to 2.6-GHz range. WiMAX also uses OFDMA for both uplink and downlink, unlike LTE, which uses OFDMA for the downlink and SC-FDMA for the uplink. Figure 1 shows a WiMAX basestation that is integral with its antenna on the tower mast.

WiMAX is basically a 3G technology, although the latest standard version, 802.16m, is considered one of ITU’s selected IMT-Advanced 4G technologies along with LTE Advanced. This more advanced version, also called WiMAX 2, has been defined as part of the ITU’s IMT-Advanced standard. It uses wider bandwidth and MIMO to achieve 1-Gbit/s data rates. GCT Semiconductor recently announced a single-chip WiMAX 2 802.16m chip that implements 4x4 MIMO (Fig. 2).

Assaf Katan, vice president of corporate development for WiMAX vendor Alvarion, indicates that the WiMAX business divides up into two major categories: carriers and enterprise. The carriers offer WiMAX broadband wireless service and/or cellular phone service. The primary U.S. carriers are Clearwire with its fixed/mobile service and Sprint Nextel with its cellular and mobile service. UQ Communications uses WiMAX in Japan.

In the enterprise space, the application is private outdoor long-range links or networks for companies or municipalities. A growing business category is radio access network (RAN) complementary applications that use WiMAX for data offloads of other wireless networks. Distributed antenna systems (DAS) also use WiMAX to multiply network capacity in large venues like stadiums.

Katan says that many developers are now building hetrogenous networks rather than starting with a macro network and filling in the gaps with other technologies as required. The hetero approach employs all relevant technologies to optimize the use of existing spectrum and to maximize capacity and data rates.

WiMAX is well entrenched worldwide, especially outside the U.S. In the U.S. the trend favors more LTE installations. The roadmap for most WiMAX companies and carriers is Time Division LTE (TD-LTE). It’s very similar to WiMAX, which is a TDD technology, unlike the standard FDD LTE that the cellular carriers are adopting. TD-LTE provides better capacity and mobility and is a relatively easy transition from WiMAX. However, the 802.16m version may never really get off the ground.

Clearwire and Sprint in the U.S. and UQ Communications in Japan are WiMAX carriers. Clearwire/Sprint has a nationwide WiMAX network with USB dongles as well as a selection of WiMAX cell phones like the HTC EVO 4G, Nexus S 4G, and Samsung Epic 4G. UQ has USB dongles and plug-in interface cards for laptops. Clearwire’s CLEAR Spot is a WiMAX hotspot that can connect up to eight Wi-Fi sources to the network (Fig. 3).

Wi-Fi: Wherever We Need It

Wi-Fi, which comprises the IEEE’s 802.11 family of wireless standards, is basically a local-area network (LAN) technology, but it’s also widely used for broadband connections. In some cases, it’s just an interim link from a PC, laptop, tablet, or smart phone to a wireless router connected to a cable or DSL Internet service. But it can also be a direct connection to a hotspot in an airport, hotel, municipal mesh, or other public place, paid or free.

Another growing alternative is the 3G/4G mobile hotspot. These wireless devices connect to a 3G/4G cellular network but also serve as portable hotspots for four or five or more Wi-Fi-enabled devices. The Novatel Wireless MiFi mobile hotspots support 3G EVDO Rev A, HSPA, and LTE. Verizon’s HTC Thunderbolt 4G LTE phone can serve as a hotspot router for up to five Wi-Fi devices.

Wi-Fi is ubiquitous. It’s built into virtually every mobile device, especially smart phones, tablets, notebooks and netbooks, and most desktop PCs. Even printers, cameras, and other devices now include a Wi-Fi radio. Most of the new Wi-Fi chips use the IEEE 802.11n standard. Most Wi-Fi transceivers are more commonly part of a combo chip like the Texas Instruments WiLink 7, which includes Wi-Fi, Bluetooth, GPS, and FM radios (Fig. 4). With its presence in most mobile devices, Wi-Fi is a no-brainer for many broadband schemes.

Finally, Wi-Fi’s use as an offload option for cellular networks is growing. This application is a recent development but a key one. The data capacity of most wireless networks, both 3G and 4G, is still limited. And with the increasing number of heavy data users accessing video, most networks can quickly run out of bandwidth when multiple users access the network simultaneously.

This either crashes or slows the network, leaving users with a bad experience that can even lead to churn, something no carrier wants. Yet building out the network with more and faster basestations is prohibitively expensive, at least in the short term.

Now most of the major carriers are attempting to divert some fast data streams to a parallel Wi-Fi network, unloading the main cellular network. This means access to a wide-ranging Wi-Fi network in areas where video and other heavy data traffic is common.

The Wi-Fi offload works like this. A consumer accesses a site with video or other data and starts to download it. If the data could overwhelm the cellular network, the data stream is seamlessly diverted to the affiliated Wi-Fi network, if available, with its inherent high speeds and separate backhaul. The user often never knows the difference.

This approach requires the smart phone, tablet, or laptop in use also to have internal Wi-Fi that has been pre-configured to deal with the service set identifier (SSID) and password automatically. Most major carriers (AT&T, T-Mobile, and Verizon) already use such a system but only in areas where data traffic demands it and a suitable Wi-Fi network affiliation has been established. For more details on offload, go to the Wireless Broadband Alliance’s Web site at www.wballiance.org and the Wi-Fi Alliance’s site at www.wi-fi.org.

One vendor with a carrier-class version of Wi-Fi, Wavion Wireless Networks of Israel, has fine-tuned and enhanced the traditional 802.11a/b/g/n radios to not only extend their range and reliability for rural broadband use but also to implement the 3G/4G offload function. Its WBSn basestation units use a unique 3x3:3 MIMO arrangement that can aggregate gigabit capacity and 450-Mbit/s speeds (Fig. 5).

These basestations also offer spatially adaptive two-way beamforming with the 802.11n standard to achieve maximum range and capacity for an outdoor Wi-Fi system. Furthermore, they use a high-gain diversely polarized (HGDP) antenna array and Wavion’s interference immunity suite to ensure solid link connections even in NLOS operation.

Independent Wireless Services

While the main telecom carriers are the primary suppliers of broadband, there are several existing and forthcoming independent broadband wireless service providers. The most visible are Clearwire and LightSquared.

Clearwire runs its popular CLEAR nationwide WiMAX service using its extensive 2.5-GHz spectrum holdings.Its network of WiMAX basestations covers more than 80 U.S. cities and is available to an estimated 130 million potential users.

The company offers home modems for fixed service as an alternative to cable TV or DSL service. It also provides USB dongles for laptops for mobile operation. Download data rates vary with range, LOS conditions, noise, and other factors, as is normal for any wireless service.

Most users experience an average of 4- to 6-Mbit/s data rates with peaks to 10 Mbits/s. Data plans are very affordable. Clearwire also just released its CLEAR Spot mobile hotspot, which connects to the CLEAR network and can service up to eight Wi-Fi connections to smart phones, laptops, tablets, or whatever.

In fact, Clearwire recently announced that it will transition from WiMAX to LTE over the coming years. The company will overlay a TDD LTE network over the existing WiMAX network and maintain both. (For more, see “Clearwire Announces Plans To Add LTE To Its Network” at www.mobiledevdesign.com.)

LightSquared is proposing another interesting system under development. The company is planning to build a nationwide LTE network that will serve not only the major cities but also most rural areas. This matches up well for the National Broadband Plan.

LighSquared will sell its services wholesale only, meaning it will not sell directly to end users but to other companies that want to establish a wireless broadband presence. Ultimate customers may also be smaller cellular suppliers who want to use LTE but cannot or will not afford the capital expenditure.

LightSquared’s approach is to use the L-band (depending upon the defining source, 1 to 2 GHz), unlike the traditional cellular bands. Its proposed operating frequencies are in the 1525- to 1559-MHz range. The biggest issue is that GPS also uses the L-band operating at 1227.6 MHz and 1575.42 MHz.

GPS satellite signals are very weak and are subject to interference from other services. As a result, there is considerable opposition to LightSquared’s future network. Most big users of GPS including commercial as well as government and military organizations oppose the LightSquared plan. The FCC is now reconsidering LightSquared’s plan to use its licensed frequencies.

The Federal Aviation Agency, which is working on its next-generation navigation and landing systems based on GPS, opposes the plan. LightSquared has proposed several solutions such as using the lower end of its spectrum to provide greater spacing in addition to using effective filters and lower power. Yet the opposition continues. The FCC has not yet decided what to do. The fate of LightSquared and its potential seems in jeopardy.

One possible solution is a recent Sprint-LightSquared agreement that provides a way for Sprint to enter the LTE market (Sprint uses WiMAX now) by using LightSquared’s wholesale services. The agreement also allows LightSquared to license unused Sprint spectrum in the 2.5-GHz bands for its network.

What About Femto Cells?

Femto cells are hand-sized basestations designed for home or businesses where indoor cellular coverage is poor. While they do facilitate faster broadband in cell phones, tablets, and laptops, they also rely on an independent broadband connection for backhaul. This is usually cable TV, DSL, or fiber. So, yes, femtos do indeed make wireless broadband possible, but they use another wired broadband connection. They are less a factor in wireless broadband than other methods.

The National Broadband Plan

In March 2010, the FCC laid out a National Broadband Plan with the goal of extending broadband to all U.S. citizens over the next 10 years. At the time of the plan, broadband adoption was estimated to be at 67% of the U.S. homes with about 24 million citizens underserved by broadband (see “Broadband For Everyone” and “Defining Broadband” at www.electronicdesign.com). The big question is what progress has been made, if any, since then.

First, the FCC is beginning to address the spectrum problem that afflicts the wireless broadband effort. The goal was to provide an additional 500 MHz of spectrum to be used in broadband with 300 MHz of it promised to mobile broadband. Some of that has already been identified.

In another effort, FCC Chairman Julius Genachowski is promoting volunteer auctions that would allow TV broadcasters (and presumably others) with extra spectrum to auction it off for broadband purposes. In a speech on March 16, 2011, Genachowski implied that the plan emphasizes the mobile aspect of broadband, as that’s clearly the trend and the future. That’s why the spectrum acquisition initiatives are so important.

The National Telecommunication and Information Agency (NTIA) also recently conducted a broadband assessment to see just what broadband services are out there and how many are using them. The organization mapped the U.S. as a way to determine who has or needs broadband and where. For example, rural areas lack suitable broadband connections. For results, go to www.broadbandmap.gov.

Progress is also being made in establishing a nationwide interoperable public safety network that would allow fire, police, EMS, military, and other public service agencies to communicate during a disaster. This complex problem is being addressed by the newly formed Emergency Response Interoperability Center (ERIC). It is a work in progress, but the plan is to have such a network built out by 2020.

Trends and Issues

The trend toward more and better wireless broadband is clear. It is here now and growing. Furthermore, there are multiple options. The technology is being used right now in existing smart phones, tablets, and other terminals. It’s maturing, stable, and affordable, and there are roadmaps to the future. Chips for dongles, routers, handsets, tablets, and laptops are abundant. The issues for growth are less technical and more related to business and regulations.

Perhaps the greatest deterrent to wireless broadband expansion is carrier rollout costs. Building the infrastructure is tremendously expensive, and it does take time. Carriers want to get the most from their 3G expansions first before going full blast with 4G. New basestations and backhaul are both essential. LTE and other major buildouts cost billions, so look for a gradual expansion.

Next, carrier data plans need to evolve to handle the new traffic. And that traffic won’t be forthcoming unless the rates are reasonable. That means more flexible plans including multiple family discounts. In the meantime, carriers will use offload and throttle some high-volume users until their buildout can handle the load. For some broadband suppliers, an appropriate business model is still needed.

Finally, a key issue affecting all carriers and vendors is spectrum. We have enough for now. But as the demand for more wireless broadband increases, as it surely will, we will face a spectrum crisis even greater than the current shortages. The FCC and NTIA will have their hands full in the next few years sorting all this out with white space decisions, government versus commercial allocations, and new auctions.


Lots of factors are influencing the growth of wireless broadband:

  • The significant growth in smart-phone sales: According to Juniper Research, smart-phone shipments are expected to reach 1 billion annually by 2016, up from 302 million in 2010.
  • The rollout of LTE and other 3G (HSPA) and 4G networks: Carriers are looking for more data income to support a continuously declining wired phone business and saturated voice cellular services.
  • The introduction of tablets and their subsequent acceptance and increased usage: Research firm In-Stat estimates that about 86% of tablet owners watch video on their devices and projects video consumption to rise to 693 billion minutes by 2015.
  • The impetus of the government’s National Broadband Plan: Expect the Federal Communications Commission to continue its efforts to extend broadband access across the nation, including rural areas.
  • The availability of other wireless systems like CLEAR from Clearwire and the forthcoming LightSquared: As other companies enter the broadband market, expect availability to increase and prices to decrease.
  • Greater flexibility and affordability of data plans from the carriers: Broadband usage will extend beyond the home and into the mobile market, cutting across demographics, and companies will respond to underserved and evolving consumer needs.
  • Increased consumer interest in mobile TV and the availability of attractive video options: One Cisco projection says that mobile data traffic, mostly video, could increase from 0.6 EB (exabytes, 1018 bytes) in 2011 to 6.3 EB in 2015.
  • The acceptance of the Internet-everywhere concept and addiction by the public: As we gain access to the Internet in more places and via more devices, mobile will become the new norm.


The key to handling the coming mobile data tsunami is greater speeds. That translates into more bandwidth. And bandwidth is limited by the carrier’s licensed spectrum as well as just plain old future availability of new spectrum. With spectrum shortage already a problem, other solutions are needed.

Shannon’s law says that data speed is proportional to bandwidth and signal-to-noise ratio (SNR), and it cannot be violated. So, we need to find a workaround.

ISCO International is working on the noise problem. Its PurePass DSP solution conditions the reverse RF link of the cellular link to improve SNR. Improving SNR in turn permits the higher-order modulation methods like 64-state quadrature amplitude modulation (QAM) to achieve the full speed it is capable of.

The reverse link is the path from the cell phone to the basestation. It is plagued by co-channel interference and adjacent channel RF power leakage. The basestation receiver sees these factors as noise. The DSP conditioning removes interfering noise and materially boosts data rates.

Steve Perlman, the Silicon Valley inventor of Rearden, has invented a new wireless system called Distirbuted Input Distributed Output (DIDO) that promises to increase data speeds for a given amount of spectrum. It essentially overcomes the problem of speed sharing as multiple users try to access the same spectrum.

As more users attempt to access a cell site or a Wi-Fi access point, the total speed is divided up among the users. The more users, the slower the connection. Perlman’s DIDO fixes this problem.

If a user attempts to access a video Web site via a cell site or access point (AP), that user gets full bandwidth and the fastest data rate as long as there are no other users. With multiple users, the data rate falls.

DIDO takes the data from the accessed Web site and process it in the cloud in a DIDO data center with complex algorithms to create a special signal that is sent to a special DIDO AP or cell site and then to the user. Even if there are multiple users, the data center processes each source and creates a special signal for each user.

The signals are all transmitted on the same channel at the same time, which are addative. Yet the user sees only the special signal, which provides the same speed as if the bandwidth was not shared.

Multiple-input multiple-output (MIMO) does something like this while processing multiple streams but can only boost data rates four or five times before it runs out of steam. DIDO is expected to produce data rate connections 10 or 100 or more times typical rates.

The DIDO system works in the cloud and requires special APs and data centers. The system is still in development, but Perlman and his crew have tested it and patents are on the way. It remains to be seen how practical it will be. It could be a real breakthrough.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.