CHIP magnet
It’s the ultimate goal of every manufacturer and the dream of every wireless user: a communications device flexible enough to work on any network. With it, users could talk or download media wherever a wireless connection is available instead of being beholden to the whims of the industry’s alphabet-soup lexicon of GSM, CDMA, iDEN, Wi-Fi and so on.
Such a scenario is possible with software-defined radio (SDR), a much-anticipated technology that offers a virtual assurance of connectivity to enterprises and an ideal technical gateway to interoperability for public-safety entities.
Hailed by some as the wireless industry’s Holy Grail, multimode SDR for the last decade has been a technology deemed to always be at least five years away from the mass market. Although SDR has been used successfully by the military, those designs have been far too expensive to be replicated in the public-safety or enterprise markets.
But this economic conundrum could change drastically in the coming months. Fabless semiconductor start-up BitWave Semiconductor has announced the development of its Softransceiver, an SDR transceiver that company officials say will make multimode, multiprotocol devices more effective — and more affordable.
Currently, multimode devices are more expensive than single-mode devices, because a separate transceiver is required to accept signals for each mode, said Al Margulies, SDR Forum executive director. “[That] takes up more space and adds more weight,” Margulies said.
Such devices also are more power-hungry than single-mode radios, said Doug Shute, BitWave CEO and co-founder. With this in mind, BitWave set the lofty goal of developing a software-defined transceiver that would cost no more to produce than a single-mode transceiver and would not increase the device’s power consumption, he said.
Using $13 million in funding raised in November 2004, BitWave has designed a CMOS RFIC chip that will transmit and receive signals from 700 MHz to 4.2 GHz and has proved that it works with the receiver and transmitter on separate chips, Shute said. Russ Cyr, BitWave’s chief marketing officer and co-founder, said 2006 will be devoted to integrating the transmitter and receiver onto a single chip and producing Softransceiver chips in volume.
If BitWave has succeeded in developing a multimode transceiver that can be produced at the same cost as a single-mode transceiver and doesn’t use more power, it has something special, Margulies said.
“Normally, a programmable chip will use more power and cost more than an ASIC chip,” said Margulies, who is not familiar with this technology. “If they’ve met that [power] goal, then they’ve got something new.”
Indeed, TechnoConcepts — another company developing a software-programmable transceiver chip this year — plans to use a silicon germanium chip for the analog-to-digital conversion that uses the power of a normal PC computer chip. Although TechnoConcepts officials note that its transceiver chip is 95% CMOS and will operate at “comparable” power levels to current transceivers, the performance needed to make SDR a reality dictates that more power will be needed.
Shute said this typical SDR architecture also requires a power-hungry digital signal processor, a combination that would not be accepted in a cell-phone market focused on limiting costs and extending battery life.
With this in mind, BitWave opted to design its solution to work in bulk digital CMOS — the least expensive silicon process on the commercial market — without using any more power than today’s single-mode transceivers. According to Shute, BitWave’s transceiver will use less than 10% of the power required by a typical SDR solution.
“We have a more conventional [cell-phone] architecture that brings the signal down to either low IF (intermediate frequency) or zero IF before the A-to-D converter goes to work on it,” he said. “So, you can have a slower speed, lower power consumption and lower cost than an A-to-D converter solution.”
John Jackson, wireless analyst for the Yankee Group, said the BitWave proposition is very compelling — if the company can deliver its promises.
“I think they have done something very innovative. … If it works as advertised, it is disruptive,” Jackson said. “If you accept that the world is moving to an era of multimodal, ubiquitous connectivity, then these guys are in an appropriate place to facilitate that transformation.”
Who are these guys?
BitWave may be a start-up, but its co-founders are not new to SDR. In the 1990s, Shute and Cyr headed Steinbrecher, a wireless company that developed the first SDR base station before being bought by Tellabs, Shute said.
But the challenges of building SDR-capable handsets are much more difficult than those associated with SDR base stations, Margulies said. With such complex equipment, the potential for costly mistakes in the development process is greatly increased, he said.
“In general, the issue is cost, power and size, when it comes to handsets,” Margulies said. “And the handsets are much more sensitive to that than base stations, vehicular radios and military radios. If you’re plugging a base station into the wall in a shelter, an extra couple of ounces, an extra couple of dollars or an extra milliwatt or two really doesn’t make the same kind of difference that it does in a handset.”
Indeed, “the most expensive real estate in the world is on a cell-phone board,” said Tanuj Raja, vice president of business development for Sandbridge Technologies, which has developed a software-programmable baseband processor that Raja believes is “complementary” to BitWave’s Softransceiver architecture.
The key development has been the rapid evolution of semiconductor manufacturing. Driven by Moore’s law, chips now can be produced with the processing power and form factor necessary to make a software-defined transceiver a reality for the mass market, Shute said.
“Ten years ago, you didn’t have .13-micron CMOS,” he added. “[Semiconductor production advancements] enabled us to economically implement a solution like this.”
Cyr said BitWave has taken each component of a transceiver and made them software programmable to handle any protocol at any frequency.
“We make it operate as a GSM [low-noise amplifier] or flip some bits and make it operate for Wi-Fi — a different frequency, a different bandwidth, a different linearity requirement,” he said. “Instead of making it fixed for one operation, we made it tunable on the piece of silicon under digital control. Then we went through the entire transceiver chain and did the same thing with every single component.”
Although complicated to test and design, the single-transceiver result will utilize much less space/power/money than two transceivers — or more.
“Now, if you wanted to add WiMAX to that, [today’s multimode] solutions would require adding a third transceiver,” Cyr said. “With us, it’s just more programming.”
First step: cell phones
“The idea is to have one phone that will work anywhere in the world that’s enabled by one transceiver chip,” Shute said.
This one-sentence company goal potentially has powerful implications for the cellular industry, which is BitWave’s first target market.
For handset vendors, a programmable transceiver means a streamlined design process that will let manufacturers focus on feature sets instead of the protocol and frequency that will be used. Cyr said some large handset makers have to build 12 different versions of every wireless phone they want to market, which lengthens the design phase and upsets normal economies of scale.
“When Motorola wants to roll out the RAZR, they literally have to bring out different versions — one for Verizon, one for Cingular, one for Orange in the U.K., one for Telesp in Brazil,” Cyr said. “Imagine now, instead of doing that, they put in one programmable device. When they ship it to Cingular, it’s the GSM version of the cell phone; when they ship it to Verizon, it’s CDMA; and when they ship it to Telesp, it’s a different frequency.”
For wireless carriers, an SDR device lets an operator market new services to customers without requiring users to buy an expensive new phone — or forcing them to sign a long-term contract in return for subsidizing the cost of a feature-rich device, Cyr said. With an SDR phone, the new service can be started with a simple software download instead of a trip to a dealer.
“Carriers are trying to raise the average revenue per user, and just offering cell-phone services isn’t doing it anymore,” he said. “They need to offer new services, but the only way to offer new services is if they can turn on new services in the device. … Now, they can have a new stream of revenue from the user, and they’ve done it through a software upgrade.”
Certainly the ability to turn on new services easily is attractive to customers, too. But perhaps the most attractive aspect of an SDR phone is that an owner could switch carriers — even those using different protocols at significantly disparate frequencies — without having to buy a new device and manually re-enter contacts and favorite sites.
“Imagine if you had to buy a new landline phone every time you switched long-distance carriers or had to buy a new TV when you changed your video provider,” Raja said. “Essentially, that’s what we do in wireless today. … But with [SDR], you wouldn’t have to.”
Multimode SDR radios also would create other flexibilities within the industry, Shute said. Mobile virtual network operators — which do not own their own spectrum — could more easily change network partners — or have multiple network partners — based solely on the best business deal available instead of having to weigh often overly burdensome technical costs associated with a transition.
In addition, carriers could forge more flexible roaming agreements with any network operator covering a given area instead of being able to negotiate only with carriers that use the same technology. For example, Cyr said a conversation could be switched from 1.9 GHz to 800 MHz in a manner that’s imperceptible to the user.
“What we see happening is that the baseband will realize that the signal is getting weak from, say, the Sprint network, and it will command our software-programmable transceiver, ‘Why don’t we change from CDMA to iDEN? Let’s look over here and see if we can find a signal that lets us continue our communications,’” Cyr said.
This capability should be especially attractive to the newly merged Sprint Nextel or the Verizon/Vodafone partnership, Shute said. Indeed, the Yankee Group’s Jackson noted that the impact of multimode SDR phones on wireless industry consolidation — for manufacturers and carriers — could be dramatic.
“This has the potential to mitigate the fragmentation that exists across operators and network deployments and modulation schemes,” Jackson said. “That sets the foundation for all sorts of new partnerships, acquisitions and tie-ups that at this point you could just begin to speculate on.”
Whether wireless carriers are going to be able to sort through the business aspects of this SDR-enabled environment may determine how quickly the technology appears in cell phones, said Ron Hickling, TechnoConcepts’ chief technology officer.
“I interpret the problem not as a technology problem … but really as a business-model problem,” he said. “If I’ve got a CDMA network, part of the way I protect my network is that [a customer’s] CDMA phone can’t go on a GSM network. The moment I create the device that can do both, I need to figure out a new business model, part of which includes exchanging traffic and billing one another. I think uncertainty about that situation is part of why there is slow movement … to doing complete SDR-type stuff.”
Beyond Cellular
Although mobile phones are BitWave’s target market, company officials are cognizant that the Softransceiver has applications outside the lucrative commercial cellular market. Cyr said BitWave already has been approached by vendors interested in developing a “universal data card” that would provide users access to EV-DO, HSDPA, Wi-Fi and WiMAX networks — an attractive notion to a mobile enterprise user that needs to be connected at all times, regardless of location.
Such a data card also would be attractive to the public-safety market, which is increasingly using data channels to enhance communications and offload traffic from its voice networks. And, of course, public safety long has been one of the most interested parties in the development of multimode SDR devices that could solve many of the technical hurdles associated with interoperability. In fact, TechnoConcepts believes public safety may be the most logical market to use software-programmable transceiver technology, Hickling said.
“I think the cellular market believes frequency agility is really nice, but it’s not really required,” Hickling said. “I don’t think that market really values it as much as a market like homeland security, where interoperability is critical.”
To date, SDR technology has not been offered at a price point public-safety organizations can afford. But a software-defined transceiver that can be produced economically can change that paradigm, Cyr said.
“You could now build a single public-safety radio that’s small in size, low in cost, yet works on all the different things that are going on out there — one at a time — simultaneously,” Cyr said.
And, while not practical for the consumer market, Cyr said a public-safety radio could be built that is not limited to terrestrial networks — a feature that has become particularly attractive after failed LMR networks hampered recovery efforts in the aftermath of Hurricane Katrina.
“You could run around doing everyday activities and be on the public and private networks,” he said. “When a disaster happens, it could search to find the operating networks, and when you find none, you could go up to the satellite.”
Although the public-safety community is perceived to be rather conservative from a technological standpoint, SDR is a new technology the sector is willing to embrace quickly, said Fred Franz, chairman of the SDR Forum’s Public-Safety Special Interest Group.
“I think [acceptance] would be a significant issue if it were a major swap out … but you’re not going to see a radio that is going to look and feel different than the ones you’re used to seeing today — it will just have additional capability,” Franz said. “So I honestly don’t think there is a major issue in the acceptance of the technology.”
In addition to handsets that could work on UHF, VHF, 800 MHz and 700 MHz, SDR radios could let public safety utilize “white spaces” — unused spectrum in the TV bands — when spectral capacity becomes an issue, such as when first responders are dealing with a major incident, according to SDR Forum’s Margulies.
But SDR’s entrance into the public-safety arena could lag well behind the commercial sector for business reasons, Margulies said. With no single large buying presence, the fragmented public-safety market is “driven by the vendors, not the customers,” he said.
“The public-safety folks that we know are interested because they see advantages, in terms of interoperability, system control and other things,” Margulies said. “But it’s a matter of getting all the requirements together and getting manufacturers to agree, and this is not a group of manufacturers that agree under normal circumstances.”
The cognitive promise
This is just one example of the many business realities that will need to be resolved as SDR finally hits the market and the industry begins working toward cognitive radio — a platform that “learns” to use spectral and network resources more efficiently based on previous experiences.
“Before you can get to cognitive radio, the first thing you need is a device that’s actually tunable because you can’t be cognitive until you know what’s out there,” Cyr said. “[Our solution] is sort of the underpinnings of what cognitive radio will be in the future — it’s the technology that enables cognitive radio.”
Everyone interviewed for this article agreed the impact that SDR and cognitive radio could have on spectrum policy and spectrum auctions throughout the world could be enormous. Whether SDR and other technological advances that optimize spectrum usage eventually commoditize the airwaves is merely a point of theoretical speculation at the moment, but that will be an assessment every wireless operator worldwide will have to make during the next several years while considering potential business-model changes.
In the meantime, BitWave officials are optimistic about the near-term outlook for the company’s software-defined transceiver.
“It not only has benefits to them [carriers], it has benefits to the user, and it has benefits to the handset maker,” Shute said. “Everybody sort of wins with this approach.”
SDR transceivers: Road map to market
June 2005: TechnoConcepts demonstrates software-defined receiver chip that processes radio signals between 450 MHz and 1.2 GHz.
December 2005: BitWave Semiconductor receives software-defined silicon for receiver and transmitter chips from its factory.
February: BitWave scheduled to demonstrate separate receiver and transmitter silicon at 3GSM in Barcelona, Spain. Similar demonstrations will be made in March at CTIA in Las Vegas.
June: BitWave plans to integrate receiver and transmitter into a single transceiver chip, known as the Softransceiver.
Summer: TechnoConcepts scheduled to have samples of its True Software Radio (TSR) transceiver chip available.
December: BitWave scheduled to ship beta versions of the Softransceiver chips to cell phone manufacturers.
Late 2006/Early 2007: TechnoConcepts scheduled to make first production run of TSR transceiver chip and deliver to customers.
2Q 2007: Cell phones with BitWave Softranceiver chips scheduled to appear on the market.