Note from the Editors: We have interviewed many executives of UWB companies and have heard their views of the potential applications of UWB. We invited Billy Brackenridge of Staccato Communications to share his view of where UWB is going.
Billy Brackenridge is a product system architect at Staccato Communications, where he helps customers build innovative products using Staccato’s Ripcord™ family of products, a single-chip, all-CMOS solution based on Certified Wireless USB. Earlier in his career, Billy helped pioneer multimedia at the Voyager Company in Santa Monica developing CD-ROM and laser disc interactive titles, and then worked as a program manager in the operating systems group at Microsoft where he worked on USB, audio and Bluetooth.
Ultrawideband (UWB) is a new technology that will enable digital wireless connections. Market researchers are predicting several hundred million UWB chips will be sold in 2008. Most of these chips will end up in the home. How will they be deployed?
I am writing this in December 2006. To date no UWB chips have been sold. Is this another ‘zero billion dollar a year’ industry? How can I be so confident in my predictions?
The answers lie in a combination of physics, business and law.
The Physics of UWB
In 2002 the FCC opened up a large portion of spectrum for use in consumer devices. The spectrum ranges from 3.1 GHz to 10.7 GHz. This is an unprecedented amount of spectrum and hence comes the name ‘ultrawideband’.
While traditional radios concentrate energy at a specific frequency and ‘tune’ to that frequency, ultrawideband radios spread energy at low power levels over a wide range of frequencies. Since there is very little energy at any one point in the spectrum, ultrawideband transmissions neither interfere with nor are blocked by conventional radio signals.
Conventional radio has not changed much since the 1930s. Transistors and later integrated circuits replaced vacuum tubes and discrete components, but the principles of radio are pretty much the same.
The change from analog to digital has completely changed how we can build radios. Ultrawideband radios are only practical using digital signal processing. UWB radios take the signal from an antenna and digitize it. The digital signal processing is basically looking for patterns of energy over a wide area of frequencies.
Digital radio can be implemented today in a single-chip special-purpose computer. In the past, radio transmitters were all about sending large amounts of power into an antenna. Today, the individual transistors on a computer chip operate at microscopic power levels.
Large amounts of transmit power mean long range transmission. In order to double the distance, you must square the amount of power. This law of physics works in our favor when we want to build a radio that will stay confined to a single room. We can build practical UWB radios that can transmit at 100 megabits per second at a range of 10 meters. These radios are limited by law to less than a millionth of a watt in any megahertz of spectrum space.
Because there is so little power at any single point in the spectrum, UWB radios don’t interfere with conventional radios. Because the signal is spread over such a wide range of frequencies, a strong signal at any given frequency will not bother a UWB radio.
The result of all this is that UWB radios are legal in the United States and can be used for consumer devices without an individual having to get a license to operate a radio. Other countries are following with similar laws.
The Business of UWB
The physics and approval from regulatory authorities are a necessary first step, but in order to make practical products there has to be a business. UWB radios have been standardized by the WiMedia Alliance. This is a non-profit association of more than 100 companies large and small that have gotten together to standardize UWB radios.
More than $250 million dollars of venture capital has been invested in startup companies promising to build WiMedia compliant devices. This doesn’t count money in larger corporations where the R&D expenses are not publicly disclosed.
WiMedia compliant devices will start to appear on the market early in 2007, and the market is expected to expand rapidly.
Cable replacement will be the first market for WiMedia, and the USB cable will be the first to be replaced. The USB Implementers Forum has released a specification called Certified Wireless USB, and several companies are building product. The first products will be "dongles" that plug into computer USB ports, and wireless USB hubs that will let you plug in peripheral devices such as printers on the other side of a room without stretching a cable across the room.
In the not too distant future, Certified Wireless USB will come built into PCs of all varieties—just as USB ports are built into all PCs today. Modern laptops have a socket under the keyboard called a PCI Express Mini Card slot. The socket is designed to support Wi-Fi and Bluetooth adapters, and provides a connector to an antenna built into the laptop lid. At Staccato Communications, we are building a UWB Mini Card where a WiMedia radio shares the antenna with Bluetooth; because the two radios do not interfere with one another, both can operate sharing a single antenna.
The traditional PC in the home won’t change much. Kids still need PCs to do their homework. PCs will get cheaper, smaller and quieter. Wi-Fi will remain as the main method for internet access in the home. The VGA cable is an unnecessary cost as the main CPU and memory is small enough to be built into the flat panel display. Printers may use Certified Wireless USB, but they might use Wi-Fi. Whichever camp produces the cheapest and easiest to use solution will win. I think Certified Wireless USB has the edge on ease of use, as everybody understands wired USB. In any case, cables on conventional computers are a thing of the past.
Once WiMedia radios are built into PCs, consumer electronics manufacturers will start embedding them in devices. Digital cameras will probably be the first consumer devices with embedded WiMedia radios. All digital cameras come with USB connectors today, but using a cable with a camera is particularly cumbersome, and it’s easy to forget to bring the cable with you when you travel.
Wireless Video Everywhere
Another relentless trend is that processors are running at lower and lower power, and this enables portable media players. Intel believes that audio and video will be played on Intel processors, and the trend is to get them to run in smaller and smaller packages that don’t require fans. Intel’s competition believes special-purpose architectures will allow lower power and lower cost video encode and playback. There are numerous examples of these sorts of chips: Texas Instruments uses the trade name DaVinci™ for its line of video processing chips; ATI (now part of AMD), Broadcom, Qualcomm and Micronas also make video-capable chips than can fit in portable players, mobile phones and TV sets.
One thing all of these media chips have in common is USB interfaces. More than three billion USB sockets have been shipped to date, and another billion are expected to ship in the next twelve months. Software drivers for USB are well understood and widely available for every family of processors. USB is not just for Intel processors and PCs anymore.
Thus the transition to Certified Wireless USB won’t require redesign of existing media processing hardware. A decade ago planners believed 1394 would be the cable of choice for video, but USB won in the marketplace. As John Lennon put it "Life is what happens to you while you're busy making other plans."
The next big issue is what becomes of the CATV cable itself? Today CATV provides a thousand video and audio channels as well as internet access all on one cable.
The first step in accessing CATV is the TV tuner. Until February 2009 when analog TV broadcast will cease, the US government requires that CATV still carry analog TV channels. You can hook your 1950s black & white TV to a modern cable, and it will work. No cable box is required.
While most TV channels are now digital, a TV tuner is still required. Several digital TV channels are packed into conventional channels that are backward compatible with standards from the 1950s.
Microsoft’s Media Center has created a market for USB TV tuners. Today these have shrunk to the size of a USB memory stick. One can attach a cable to one of these tuners and watch TV on a laptop. The wholesale cost of parts necessary to make a modern USB TV tuner are about $25, and the resulting product usually sells for less than $100. Media center supports up to four TV tuners, and will support multiple displays, picture in picture, or combinations thereof.
USB tuners don’t have to be associated only with Intel processors or Microsoft software. Nearly every video decoder chip used in set top boxes and TV sets already has a USB interface. With the addition of Certified Wireless USB, it is possible to separate the TV tuner functionality from the viewing experience. There is no need for the TV tuner to be in the same box as the display or set top box.
For example, the TV tuner could be in a box on the wall near where the TV cable enters a room. Certified Wireless USB means you don’t have to tear out a wall or route cable under the rug. You could easily send several encoded TV streams across a room with Certified Wireless USB. One tuner could service several sets or a Media Center with several displays.
This scenario sounds very sensible, but this isn’t how it was supposed to be. A decade ago planners believed there would be one video decoder chip in the set top box. The idea was that TV sets would be ‘dumb glass’ and connected to the set top box by some sort of digital or analog cable. Then MPEG decoding got very inexpensive and consumers demanded higher and higher TV resolutions. Cables that carry raw digital video (DVI, HDMI) can cost between $30 and upwards of $200. Digital video cables are significantly more expensive than USB cables.
But the whole point is to get rid of the cable. The highest quality compressed video requires about 20 Mbps, but raw video can require up to 6Gbps. Some people are promising proprietary ultrawideband systems that can carry raw video at these rates, but Certified Wireless USB can carry compressed video for a lot less money.
Mobile Phones in the Home
Bluetooth is used for communications in mobile phones and other battery powered devices. Since it consumes very little battery power, Bluetooth is a good fit for these devices. But Bluetooth is limited in its maximum bit rate and is inadequate for transfer of media files. Attempts to build MP3 players with Bluetooth have proved unsatisfactory. Video is completely out of the question.
The Bluetooth SIG has adopted the WiMedia radio as the basis for their next standard, which they are calling Bluetooth 3.0. With WiMedia as part of “high speed Bluetooth,” consumers should be able to share media files between battery powered devices nearly instantaneously.
Intel has claimed that Certified Wireless USB will be able to download a ‘full length motion picture’ in 80 seconds. The Bluetooth SIG has not made similar claims, but the prospect of keeping videos or motion pictures on phones or iPod like devices is a real possibility. Consumers will demand the ability to share video between phones and PCs, TV sets or other phones in the home.
Bluetooth 3.0 addresses three major concerns of mobile phone operators: handset battery life, cost and efficiency of cell operation.
Because UWB transfers files faster than any other radio technology to date, it consumes far less battery power to transfer a given number of bits. As consumers demand media files on their phones, UWB is the only practical wireless technology for the next decade.
Nokia recently announced a new technology called Wibree, which complements UWB nicely in that it is power efficient for keeping connections open or transmitting a few bits. We expect to see Wibree, conventional Bluetooth and WiMedia used synergistically in Bluetooth 3.0. Wibree provides an extremely low power standby mode, and WiMedia provides an extremely power efficient way to transmit large media files.
This leaves ‘efficiency of cellular operation’ as the major factor, and is part of a broader issue of potential interference between UWB and wide-area communications.
Ultrawideband in general and WiMedia in specific were designed not to interfere with conventional radio transmission, but extremely weak signals are susceptible to even low levels of noise. By the time a signal travels from the mobile phone operator’s tower and through the walls of your home, it is very weak and subject to interference even from very weak signals. In most cases simply moving a device a few feet closer to a window will clear up interference, but frequently radios will be collocated in the same phone or laptop. There are many use cases where two devices will be near one another. For example you may download music from the phone and transfer it ‘on the fly’ to your iPod.
What this means for the consumer is that it should be possible for a battery-powered media player to download content just by being near a mobile phone. This isn’t easy. The radios for handset to cell tower communication and other wireless communication systems in the home must all operate at the same time without interfering with one another.
Today ultrawideband, Wi-Fi and mobile phones operate in separate frequency ranges and don’t interfere with one another, but spectrum is limited and we are building radios into more and more products. As consumer demand for bandwidth increases, UWB is the most promising technology that can meet the demand in a non-interfering manner.
We expect to see WiMedia systems built into mobile phones, set top boxes, automobiles, laptops, cameras, media players and general appliances. These devices will be mains powered and battery powered. As wireless becomes ubiquitous, it is essential that steps are taken such that devices don’t interfere with one another. The consumer shouldn’t care what is going on ‘under the hood’, but will demand that appliances ‘just work’. The WiMedia Alliance is taking the necessary actions today to make sure our radios scale for the future and ‘play well with others’.
The WiMedia Alliance is developing a standard called WiNet for Internet protocols over WiMedia. Today Wi-Fi is the dominant wireless technology for carrying internet protocols. WiNet differs from Wi-Fi in several significant respects.
First, WiNet uses modern hardware encryption, and encryption is turned on by default. While the newest versions of Wi-Fi use the same encryption technique, millions of Wi-Fi access points are deployed that operate in the clear and represent a continuing security risk.
Second, WiNet can communicate with several network access points and form peer-to-peer connections with several nearby devices at the same time. This may seem rather obscure, but Microsoft’s new Vista operating system is supporting new modes for PC-to-PC sharing that can exploit this feature; for example, you could give a PowerPoint presentation without a projector and the presentation would appear on everybody’s laptop screen. WiNet can provide a fast peer–to-peer connection so that screen images can be shared over the wireless link.
Ad hoc networking is a powerful new concept. The idea behind ad hoc networking is that devices find each other and assign themselves ‘ad hoc’ addresses without relying on a network administrator or centralized server. With secure ad hoc networking, devices establish a secure relationship so that they can send encrypted messages that others can’t decipher.
WiNet’s short range and high bandwidth make it ideal for secure ad hoc networking, but the experience of the internet has taught us that networking prospers best with open standards. China, Japan and Korea are pushing IPv6 as a matter of national policy. Rightly or wrongly the US is perceived as ‘dominating the internet’ and restricting Asian countries from their ‘fair share’ of IPv4 internet addresses.
Microsoft listened to their customers and implemented secure ad hoc networking based on IPv6 in Vista. WiNet will exploit IPv6 secure ad hoc networking. WiNet equipped media players will be able to exchange media files much faster than Wi-Fi. As WiMedia moves into more consumer devices and secure peer to peer ad hoc networking becomes more important WiMedia will gain market share over Wi-Fi.
We can’t predict all the ways WiMedia ultrawideband radios will be employed in the home. As with all new technologies, inventors come up with new applications that were totally unanticipated by the people building the technology in the first place. We are attempting to make WiMedia as flexible as possible so that people may find new and unexpected uses of the technology.