Broadband Advisor

A long-term battle is brewing between two emerging high-speed wireless technologies, WiMax and Long Term Evolution (LTE). Each would more than quadruple existing wireless wide-area access speeds for users.

The two technologies are somewhat alike in the way they transmit signals and even in their network speeds. The meaningful differences have more to do with politics specifically, which carriers will offer which technology, as in the recent skirmish between backers of [[ArtId:1019495862|Blu-ray and HD video|new]].

In this coming wireless war, one technology won't necessarily obliterate the other, but analysts believe LTE will have a tremendous upper hand over WiMax in coming years, primarily because carriers on the GSM standard (Global System for Mobile communications) predominate around the globe and will use LTE as their upgrade pathway. GSM is the most popular mobile communications standard.

LTE will dominate, analysts believe, despite the recent attention generated by plans for a [[ArtId:1596120987|joint venture|new]] between Sprint Nextel and Clearwire for a national WiMax network that is expected to reach 120 million to 140 million people in the US by the end of 2010. Even though Sprint officials admit that's an ambitious goal, they believe they have a clear time-to-market advantage over LTE in the US, by perhaps a year or more.

So far, AT&T and Verizon Wireless, the two biggest wireless carriers in the US, have stated plans to adopt LTE, with major rollouts planned for 2011 or 2012. And, in a surprise to many, even Sprint has not ruled out building LTE and is not prevented from doing so by its joint venture plans with Clearwire and several major investors including Google, Intel and three cable companies.

"WiMax and LTE are directly comparable in terms of what they do, and it's very likely LTE will have a significant global advantage over WiMax in the long term," said Craig Mathias, an analyst at The Farpoint Group and columnist. "But that doesn't mean WiMax is toast or won't survive, although I'm not even sure of [Sprint's] expected time-to-market advantage when we talk about critical mass penetration.

"LTE is the natural upgrade path for GSM, and that leads me to conclude that LTE will be one tough cookie for WiMax to beat," he added.

The GSM family will account for fully 89 per cent of the global market in 2011, according to Gartner. In the US, AT&T is a GSM provider, along with T-Mobile, which many believe will eventually announce intentions to support LTE.

"WiMax drives the hype for 4G [fourth-generation wireless technology], but LTE will be the dominant standard," Gartner analyst Phillip Redman wrote in an April report.

The players

In addition to the carriers, there are standards groups and manufacturers driving the two technologies. So far in the US, Sprint and Clearwire are aligned behind WiMax, while Verizon Wireless and AT&T are behind LTE.

Still, as of now, LTE is not even a set standard. However, that status might be conferred on LTE in the last half of this year by a group of vendors calling itself the 3GPP (Third Generation Partnership Project), Redman said.

LTE is in trials in the US, Europe and China, according to a Verizon spokesman, who refused to divulge the US location or any other details.

The IEEE (Institute of Electrical and Electronics Engineers) picked 802.16e as its mobile WiMax standard in late 2005, and that is the standard Sprint began deploying last fall in three markets -- Washington, Baltimore and Chicago -- in what it has begun calling a "soft launch," with its employees as users. The carrier still has not officially announced a commercial service timetable.

---PB---

A group called the WiMax Forum, made up of more than 500 vendors and other members, has established itself as an authority to certify WiMax gear. In April, it announced the [[xref:http://www.wimaxforum.org/news/pr/view?item_key=59390fb727bfa15b5b8d11bf9341b2b1176099f8|certification]] of eight WiMax products for use in the 2.3 GHz band, with products in the 2.5 GHz band expected to be certified later this year.

As for equipment manufacturers, Intel has invested billions of dollars in WiMax research and chip sets and showed off conceptual mobile Internet devices at the Consumer Electronics Show in January. Motorola, among a long list of manufacturers, has been a strong proponent of Sprint's WiMax initiative, but it also will be developing LTE technology.

The speeds

How will LTE and WiMax perform? The big technology focus of both approaches has been on speed, or throughput, to the user. Theoretical maximums for WiMax are listed in textbooks at 70Mbit/sec., while an AT&T official boasted LTE will provide speed of 100Mbit/sec.

Such speeds would greatly enhance video transmissions and online gaming. For business users, they could offer lightning-fast access to enormous corporate data stores sent over encrypted channels. Some futurists believe both technologies could bring about wireless video phone calls, like the ones Dick Tracy receives on his watch. Today, by comparison, videoconferencing from a wired desktop is still relatively expensive and not always reliable without a high-speed wired connection.

The speeds expected by both LTE and WiMax are hard to nail down primarily because the technologies are just rolling out. Many factors will be taken into consideration, including whether a carrier plans to send the signals over a wireless channel that is 40 MHz in width, double the standard 20 MHz channel, noted analyst Philip Solis of ABI Research.

Speed to an end user is also dependent on how many users are connected to a cell tower, how far away they are, what frequency is used, the processing power of the user's device, and other factors.

A Verizon spokesman wouldn't offer any predictions on what speeds LTE could deliver. "I'm not saying anything on that," said the spokesman, Jeffrey Nelson. "Not theoretical or advertised speeds or what's proved out in the real world."

At Sprint, a spokesman said conservative estimates of the WiMax downlink speeds will be 2Mbit/sec. to 4Mbit/sec. on average on the Xohm WiMax network that Sprint is building, with 10Mbit/sec. peak downlink speeds. For uplinks, the speeds will average 1Mbit/sec. to 2Mbit/sec, depending on the processing power of the user's device.

The same, yet different

There are some notable technology differences between LTE and WiMax, but analysts said both approaches have much in common.

For example, both technologies provide the same approach for downlinks, and both have Multiple Input Multiple Output (MIMO), which means that information is sent over two or more antennas from a single cell site to improve reception. In tough transmission locations, such as a dense downtown area, MIMO could be a relatively inexpensive means of improving reception to users.

The downlinks from the cell tower to the end user in both LTE and WiMax are enhanced with OFDM (Orthogonal Frequency Division Multiplexing), a technology that supports sustained video and multimedia transmissions and is already being deployed in some non-LTE and -WiMax networks. It works by splitting up signals among multiple narrow frequencies, with bits of data sent at once in parallel. Needless to say, it is complex technology that will require sophisticated base stations, an added expense even for those carriers that see LTE as an upgrade path to GSM, analysts said.

---PB---

"Although LTE is on the GSM track, it really requires new equipment" at base stations, said Lisa Pierce, an analyst at Forrester Research. That means a substantial investment is in store for AT&T and also for Verizon, which has focused on CDMA (Code Division Multiple Access), but partners with Vodafone, a GSM carrier favoring LTE. The cost of a national WiMax network will be billions of dollars, and an expenditure of that much money would not have been possible without the Clearwire-Sprint joint venture.

Uplinks from the user to the cell tower will probably be different in the two technologies. OFDM will be used in WiMax, but a technology called SC-FDMA (Single Carrier-Frequency Division Multiple Access) will be used in LTE, Solis said. SC-FDMA is theoretically designed to work more efficiently with lower-power end-user devices than OFDM is.

Both technologies will be IP-based, which will enable quality-of-service technologies to be applied, although it is not clear whether carriers will offer guarantees of service to business users. The big push for WiMax has been with consumer-based devices so far, and service guarantee considerations have been secondary.

"How much differs between LTE and WiMax remains to be seen, since LTE is not standardized yet," Pierce said. LTE is designed to turn voice and data traffic into packets, which bodes well for unified communications applications, she said. In theory, WiMax will also be able to support voice, but whether it is used for voice "remains to be seen."

The spectrum difference in the US

Both Verizon and AT&T picked up 700 MHz of spectrum in the recent Federal Communications Commission auction, and both carriers said the spectrum will be used for LTE. Meanwhile, Sprint had already been expecting to use its generous holdings of 2.5 GHz spectrum for WiMax. And the Sprint-Clearwire joint venture expects to control about 80 per cent of that spectrum for the WiMax rollout in the US.

Since two different bands of spectrum will be employed, WiMax and LTE transmissions will have different physical properties. And that, in turn, will influence the costs of building base stations and the expensive gear inside.

LTE can run on a variety of spectrums, but the lower 700 MHz frequency will provide greater range and better in-building penetration than a higher frequency would, Pierce said. Still, a lower frequency means that boosting the speed of a transmission might require technologies such as compression and the ability to bond channels together to improve performance.

In comparison, since WiMax will use 2.5 GHz spectrum, bits won't be able to travel as far as they would at a lower frequency. And that suggests that WiMax providers will need to "blanket" service areas with base stations to avoid attenuation, which is the loss of a bit's strength as it travels through air or another medium, Pierce said.

---PB---

Spectrum properties and the physical layout of cell towers and base stations are ongoing concerns of the carriers. At AT&T, engineers are moving forward with LTE after carefully testing WiMax, which is "very good technology," said AT&T spokesman Mark Siegel. "It appears to be very good for fixed mobility, such as a neighborhood or shopping mall, but it's not clear yet how it will perform over wide distances as people are mobile," he said.

Motorola, Intel and others demonstrated mobile WiMax at the CES and the CTIA conferences earlier this year in Las Vegas, driving cars through busy city streets to show connections to WiMax antennas installed for the tests. The WiMax handoff of signals from tower to tower worked the majority of times on two separate runs conducted, but failed at least once on each run. Last year, a boat on the Chicago River made smooth handoffs as it cruised along, allowing video streaming to laptops and handheld prototype devices.

Siegel said AT&T is convinced it can find cost efficiencies with LTE as an upgrade to GSM, allowing it to use its current infrastructure of 48,000 towers and related base station equipment nationally. "By comparison, WiMax has to start from ground zero," he said.

Sprint disputes that characterization, noting it has long held the 2.5 GHz spectrum being used for WiMax, among other basics, including the rights to use its existing cell towers.

The bottom line

For end users, the current debate over WiMax vs. LTE is largely theoretical but is nonetheless important. Technology investors are the most interested now, because billions of dollars being invested today will have clear implications for millions of users in, perhaps, three years.

Analysts see a clear dominance by LTE in a few years. Among other things, it could support global roaming for users of high-speed wireless devices, since so many carriers are bound to adopt it. However, that won't serve every user or every company, Redman noted.

"The bottom line is that one technology will not cover all the user needs for home, office, local or international services," Redman said. "It is still going to be a combination of technologies and developers. WiMax may be one of those, but LTE will predominate."

For users, technology competition should always be considered beneficial, noted Jack Gold, an analyst at J.Gold Associates. "One thing is certain," he said. "The new Clearwire joint venture will spur the other guys to get their act together and get LTE out in the field."