Gearing up for 700 MHz

Wireless Review, 03.31.08; Joan Engebretson
http://telephonyonline.com/wireless/news/telecom_gearing_mhz/

With the dust settling from this year's 700 MHz auction, the winners will begin to formalize their plans for that spectrum — and vendors already are gearing up to develop equipment for deployment by 2010.

Both Verizon, which won the coveted C Block licenses covering nearly all the U.S., and AT&T, which won a large number of licenses in the narrower B Block, appear certain to use their new spectrum to deliver mobile broadband services that could support high-speed data and voice over IP.

Sandip Mukerjee, vice president of wireless portfolio and strategy for Alcatel-Lucent, is particularly enthusiastic about mobile broadband. “From our primary market research, this is where we see a willingness to pay,” he said. “Be it mass-market consumer or enterprise, this is what makes the most business sense.”

The U.S. is ahead of the curve in freeing up 700 MHz spectrum by requiring UHF broadcasters to vacate the spectrum by 2009. That will give U.S. operators a rare opportunity to lead the rest of the world in their deployments, Mukerjee said.

U.S. operators, he said, “will get the opportunity to redefine traditional [average revenue per user], where part comes from end-user paid services such as voice, messaging and data access; it also ushers in non-user paid revenues from things like advertising, wholesale capacity sales and open-access devices and applications.” Noting that ARPU has been declining in recent years, Mukerjee said, “This should drive it up.”

Broadband mobile services, Mukerjee added, “will be video-rich and will be based on social networking and productivity gains from the enterprise segment.”

Because 700 MHz networks will operate at a relatively low frequency, they will provide excellent propagation characteristics, enabling base stations to cover a larger area than they would in a network operating at a higher frequency.

“Coverage will be excellent, and in-building penetration will be really good,” said Fred Wright, senior vice president of cellular networks and WiMAX for Motorola.

At least three different 4G wireless technologies have been proposed for mobile broadband use at 700 MHz. These include WiMAX; long-term evolution (LTE), which was developed by the same companies that built GSM cellular networks and equipment; and ultra mobile broadband (UMB), which has its roots in the CDMA development camp headed by Qualcomm. All three technologies appear capable of supporting mobile broadband connectivity at speeds in the tens of megabits.

At first glance, WiMAX might appear to have a time-to-market advantage over the other alternatives, as Sprint and others already are deploying it in the 2.5 GHz band. But adopting the technology for use at 700 MHz poses some significant challenges.

The 2.5 GHz spectrum band is somewhat unique in that it is unpaired spectrum, making it well suited for supporting time division duplexing (TDD), in which a single channel carries both upstream and downstream traffic — the approach used for WiMAX. But most cellular spectrum, including all but one of the five blocks in the recent 700 MHz auction, is paired. Paired spectrum lends itself well to frequency division duplexing (FDD), in which the upstream and downstream signals are carried on separate channels.

The WiMAX Forum has announced plans to develop an FDD version of WiMAX, but that will take time — and that's not the only area where WiMAX is being tweaked. The forum also is working on improving the uplink performance of WiMAX, which, according to Wright, is not as good as that of LTE.

“WiMAX is two and a half years ahead of LTE,” Wright said. “But by the time you modify WiMAX standards to improve the uplink performance and get a commercialized product, it will be neck and neck with LTE.”

Verizon Wireless and AT&T Wireless have already indicated plans to use LTE. That reality, coupled with GSM's worldwide predominance, could help ensure LTE's position as the preferred 700 MHz technology.

“The potential licensees are concerned about volume and scale and making sure they select a technology that others will likely deploy so they don't end up in a situation that would cause them to have a high infrastructure and device cost and few selections in terms of the vendors they get to deal with,” said Wright.

That logic favors LTE over WiMAX and UMB, added Wright, who questioned how much demand there could be for the FDD version of WiMAX. “It all comes back to spectrum allocation,” he said. “Around the world, WiMAX is generally deployed on spectrum that is well-suited for TDD.”

Wright added that Motorola expects to leverage the work it already has done with WiMAX in developing its LTE solutions. “We're expecting as a development company that about 75% to 80% of WiMAX software is reusable for LTE,” he said. “At this point we're focusing everything on LTE. I don't think UMB will get global traction, but we're ready to build it if people ask for it.”

Another consideration is that it will take a while for network operators to deploy 700 MHz infrastructure nationwide, which means operators likely will seek dual-mode devices supporting voice and data connectivity on 3G or 4G networks. Here, again, economies of scale will come into play.

“You'll see the device community working on dual-mode UMTS/LTE devices before UMTS/WiMAX,” said Peter Jarich, research director for Current Analysis.

Mobile broadband technologies such as LTE and WiMAX pose a unique set of opportunities and challenges at 700 MHz — particularly in the C Block, which at 22 MHz is the widest block of spectrum in the recent auction.

One potential advantage of WiMAX, LTE and UMB is that all three will support wider carrier channels compared with earlier-generation technologies. WiMAX today uses 8 MHz of bandwidth, and LTE and UMB can support channels up to 20 MHz wide. In comparison, CDMA uses 1.25 MHz channels, and wideband CDMA uses 5 MHz channels.

Newer technologies also will offer more flexibility in channel width. LTE, for example, can support channels in increments of 1.25 MHz.

Although all 700 MHz bands are appropriate for mobile broadband, the C Block will enable the highest data rates, said Danny Locklear, director of wireless network product marketing for Nortel Networks. By using 4G technologies at the wider channel bandwidths that the C Block can support, Locklear said: “You could offer services that would be more conducive to heavy video. The same thing could be done in 3G but not as cost-effectively.”

When used in the 22 MHz C Block, the broader carriers supported by 4G technologies will generate spectral efficiencies, Mukerjee said. In narrower spectrum blocks such as the A, B and E blocks, he said, “You won't get the spectral efficiency gains. There the current 3G and next-generation technologies begin to approach each other. The difference becomes more pronounced at broader bands.”

If 700 MHz winners devote a large portion of their spectrum to broader carriers, however, it will reduce the number of carriers and capacity in each cell site, which could require that cell sites be located just a few blocks apart.

“When you end up with small cells in urban areas, you can get self-interference between the cell sites themselves,” said Tom Flak, senior vice president of company operations for Soma Networks. Soma offers pre-standard WiMAX equipment operating at 700 MHz, which has been deployed by some small U.S. telcos that won spectrum in a previous auction in the lower 700 MHz band.

Flak believes developers will be able to address self-interference through careful antenna design but said that “it's an extra complication — no one has built an urban high-speed network yet.”

Another potential complication relates to multiple input/multiple output (MIMO) technology, which is used with WiMAX, LTE and UMB. MIMO increases spectral efficiency by using multiple antennas.

But as Gee Rittenhouse, vice president of research for Bell Labs explained, “To support multiple channels you have to be able to distinguish the signals coming from the various antennas. This is normally done through diversity — by separating the antennas more than a few wavelengths apart.”

But achieving that separation could be challenging at 700 MHz because the wavelengths are considerably longer than at higher frequencies.

Rittenhouse doesn't expect this to be a serious concern, however. “Diversity isn't a step function, where either you have it or you don't,” he said. “It's a continuum. If you don't have full spacing, there's not quite as much diversity, but that's not necessarily an issue.” Other possible solutions could involve using polar diversity instead of antenna diversity or increasing signal-processing power.

“The problem is complex but not insurmountable,” Rittenhouse said.

Outside the C Block and apart from the major operators, the technology path that spectrum winners will pursue is likely to be a mixed bag. Policy-makers will have to go back to the drawing board to determine the outcome for the D Block. That 10 MHz span was envisioned as a shared private and public safety network, but it did not receive the minimum bid set by the FCC. The 6 MHz E Block, meanwhile, may be a special case. Most of that spectrum was won by either Qualcomm or Frontier Wireless, a unit of direct broadcast satellite company EchoStar. Qualcomm is expected to use the spectrum to augment its MediaFLO mobile broadcast service, and EchoStar is likely to follow a similar strategy.

That leaves the A and B blocks, both of which include 12 MHz of spectrum and are based on smaller geographic territories than the C Block. Aside from AT&T and Verizon, which won some A Block licenses, most of this spectrum landed in the hands of smaller incumbent mobile operators — and whether they will deploy 3G or 4G technologies is a matter of debate.

“They could use 3G today to enhance their existing networks by providing more spectrum,” Nortel's Locklear said. “Or they could … deploy what's being used by the larger carriers in order to be their roaming partners.”

Small incumbent wireless carriers receive substantial revenues from roaming agreements with major national carriers, and they will see new roaming opportunities as major carriers deploy 4G networks. But Mukerjee believes the smaller winners of the A and B blocks are likely to deploy 3G initially because those blocks are narrower, limiting the data rates they can offer and making the business case for 4G less compelling than at the C Block.

Stakeholders are reluctant to forecast what service providers will spend on 700 MHz deployments — in part because strong take rates could erode the cost advantage provided by the spectrum's excellent propagation characteristics.

Andrew Seybold, founder and principal of the consulting firm that bears his name, estimated that about 25,000 cell sites would be needed to cover 75% of the country at 700 MHz. “To cover the same area on a 2.5 GHz WiMAX network, you would need 65,000 cell sites,” he said.

That means a 700 MHz network could be faster to deploy and cost less, at least initially. But as Seybold pointed out, “cell sites are closer together at 2.5 GHz, which means you get more network capacity.”

As traffic at 700 MHz builds, the only way to get additional capacity is to start building more cell sites closer together, which eventually could erode the 700 MHz cost advantage — in populous areas, at least.

The investment required to build 700 MHz networks also will depend on the extent to which incumbent cellular operators will be able to leverage their existing network investment. Electronics typically represent 35% to 40% of the cost of a network buildout, Wright said. “The rest is land, building and towers, and for incumbent operators, one of the advantages they have is they have existing land, building and towers,” he said.

Certain factors will tend to make 4G deployments at 700 MHz less costly than earlier-generation deployments, Wright added. “The electronics compared to CDMA will be a whole lot less,” he said. “The electronics we use in WiMAX and LTE are a fraction of the size. You get a whole lot more [millions of instructions per second] and memory on much smaller semiconductors than you did in the past. They dissipate less heat so they use more compact boxes, which will significantly simplify size and the cost of deployment.”

Where network operators may see their deployment costs rise, however, is in the backhaul network. “It's not uncommon to have a cell site with one or two T-1s to handle all the traffic,” Wright said. “In broadband systems, it will take seven to 10 T-1s to provide a minimum amount of backhaul.” Overall, he predicted that “the backhaul cost will be significant in the buildout of next-generation networks, and carriers will be looking at different technologies.”

U.S. operators traditionally have made minimal use of wireless backhaul, but that could change, Wright said. He foresees increased interest in wireless backhaul as U.S. operators seek creative solutions to their new backhaul challenges.

In the core network, IP will underlie mobile broadband networks, regardless of whether they are based on LTE, UMB or WiMAX. As a result, carriers that have invested heavily to migrate their existing core networks to IP will be rewarded, Alcatel-Lucent's Mukerjee said. “If your underlying core network is IP, you have a definite cost and service-delivery advantage that cannot be easily replicated,” he said.

Longer term, network operators deploying service at 700 MHz also may benefit from increased economies of scale as other countries open the spectrum for mobile broadband. Although few countries have announced specific plans, Seybold noted, “There is a lot of momentum in the U.N. and the ITU.”

Copyright 2008 SOMA Networks, Inc. SOMA Networks, SOMA, SOMAport, SoftAir, and the star and circle logo are trademarks of SOMA Networks, Inc. U.S. and Foreign Patents Pending. All other trademarks belong to their respective owners. IMPORTANT NOTICE - ACCESS AND USE OF THIS SITE

 
 

Copyright © 2008 SOMA
Networks, Inc.
All Rights Reserved