(top gray line)
CSC (logo) OverviewCharterFAQs
LEF HomeNews & EventsProgramsthe LEFBookshelf
(spacer image) (spacer image) (spacer image)
The Leading Edge Forum
News
(spacer image)
Events


Rich Stillman reports from COMDEX

Wednesday, November 14, 2000

Contents:
Wireless standards explained
Bluetooth explained
Timing modulation: a new approach to wireless communication
Wireless security
A few words about the Luxor Hotel

Las Vegas, NV -- Comdex is a computer show, isn't it? Well, I haven't been hearing much about computers today; at least not traditional keyboard/screen/Windows based computers. It's been a serious day of tutorials on cellular and radio technology, and a look into the near and far future of devices that truly blur the line between mobile phones and computers. As clichéd as it sounds, the development of these devices will truly change everything from the way we work and buy things to the way we talk to our kids. It's already changed the way kids talk to each other.

Since engineers were running the presentations today, much of the discussion did not involve the use of actual words. CDMA, TDMA, 1xRTT, SIM, 2.5G, WPKI - sometimes it was reassuring just to see a vowel. But the big issues here are more important than the alphabet soup and can be understood without digging too deeply into jargon. With the vast depth of knowledge gained by a full six hours of experience listening to wireless engineers talk, I'll try to translate the important stuff into English.

Where to start? The best place is probably with the digital wireless standards that will become available to all mobile phone users over the next one to four years.

2G, 2.5G and 3G wireless
The source of most of the excitement and controversy in the mobile world these days is the upgrade of the mobile phone infrastructure to allow more bandwidth and a greater variety of services. The steps in this process are described in generations.

First generation, or 1G, is traditional analog cellular and pretty much for voice only. 2G, which is widely implemented in the US and almost universal in the rest of the world, is a digital standard for voice that allows low-speed data transfer. Without getting too technical (I couldn't if I wanted to), 2G uses circuit-switching which more or less sends the information in a continuous stream like a radio broadcast.

Switching technology bears some explaining. There are only two types of switching currently in use in phone systems worldwide. Circuit switching, the older standard, works OK for voice, but it tends to lose small amounts of information along the way. These dropouts are not a big problem for voice, but they cause havoc with data. As a result, data rates on 2G systems are limited to about 28.8 kbaud, the speed of a conventional computer modem five years ago. This makes 2G fast enough for simple messaging and text-only information services, but not much else.

Packet switching is the basis of third generation, or 3G, mobile phone systems. Packet switching treats information the way the Internet and other computer networks treat data. The data is digitized, broken into small chunks called packets, surrounded it with information that allows it to be verified by the receiver, and sent off. The receiver gets the packets, verifies them, and reassembles the data. If it happens to be voice, the signal is translated back into sound. Packet switching allows a much cleaner signal, higher data rates initially up to 144 Kbaud (up to 2 Mbaud in a few years) and the ability to carry more conversations in the same amount of radio bandwidth.

The goal of all phone systems worldwide is to roll out 3G as quickly as possible. The main obstacle to that in most of the world is cost. European phone companies have spent fortunes just to buy radio frequencies to implement their 3G systems, and they can't make back a penny, or a pfennig, until they spend another fortune and a year or two to roll out the transmitters and infrastructure to make it work. Even then they'll be hard pressed to charge more than they do for their existing 2G systems until they can offer new services that take advantage of the new technology.

But the companies that can develop the new services, and the content they'll deliver, are waiting for the systems to go online so they can get an idea of what people want before they spend money on development. Start to see the problem? Phone companies are literally betting the company on investments that won't start to pay off for a couple of years at least.

In the US, all of the above applies, but things get even more interesting. Unlike Europe, where standards have been imposed by government, the US has no packet switching standard. Three distinct and incompatible systems are being used by different phone companies for digital service. None are compatible with the European standard, called GSM 1800. In addition to making things very confusing for consumers, this lack of standardization means that US phone companies are spending money to build three parallel phone systems, compared to Europe's one.

Money isn't the only problem. These independent phone networks cover different areas of the country, so a digital phone that works in one part of a city may not work in another. And different systems require different phones, causing the number of phone models to explode, again making life difficult for consumers. The end result of this is that the US will be about a year behind the Europeans in rolling out 3G to the mainstream phone market.

As one presenter’s slide said in foot-high letters, the benefits of 3G are HUGH! I think he meant HUGE, but he was an engineer so HUGH could have been another acronym. Treating voice as just another kind of data means that far more traffic can be carried in the same amount of bandwidth. This is important because bandwidth is the scarcest resource for mobile communications. The added data speed enables applications like streaming audio and graphical Web browsing, as well as new services that make no sense on a wired phone or computer but are very valuable on a mobile device. These include services as diverse as interactive gaming and direction finding, to downloading songs to use as ring tones. One of the more unique ideas is called software-defined radio. The concept is that your mobile device is a blank slate on which you download applications. If you want it to be an MP3 player, you download MP3 software and then start streaming the audio from the network. Need to check your calendar? Throw away the MP3 software, download a day-timer application, then access your network-based calendar. Voice phone or instant messaging? Download that program, and use it with your network-based address book. The end result is a much smaller, simpler and cheaper device and far more portability. Combined with authentication services, in fact, you should be able to log on to any mobile phone and instantly personalize it with your own identity and access to your own data.

Other services enabled by 3G, many of which rely on location information and always-on connection, are mentioned in yesterday’s dispatch.

How to charge for these services is an open question that deserves more space than I can give it here. But a story told by one of the presenters speaks volumes about how little idea anyone has about what constitutes a workable business model. She described a content developer sitting down with a wireless portal provider in Europe to negotiate a deal. The developer’s first question was, “Do I pay you, or do you pay me?”

2.5G: an interim solution
What to do while waiting for 3G? If you’re a phone company, you take half a step toward the all-digital, all-packet world by installing a 2.5G network. As the name implies, 2.5G implements half of the 3G architecture by using packet switching for data. However, it uses 2G-style circuit switching for voice. 2.5G offers tremendous benefits for everybody. It is much cheaper to implement than a full 3G network, and it gives content providers a high-speed data platform to start experimenting with and developing the services that will be necessary to make 3G a success. Consumers can continue to use 2G phones for voice, or buy 2.5G phones to take advantage of higher 64K to 144K speeds for data services, unlike 3G which will only work if consumers buy completely new handsets. But there is one major drawback to 2.5G.

Remember that bandwidth is the most limited resource for phone services. Since 2.5G uses the same circuit-switching protocol for voice, it requires the same amount of voice bandwidth as 2G. Data services require additional bandwidth layered on top of voice. The more speed that’s built into the data network, the more bandwidth it requires. Phone companies, most of whom are still losing money on data services, are reluctant to give up voice bandwidth for new services they still consider experimental. So 2.5G, which should be available in parts of the US within six months and is already rolling out in Europe, is purely an interim solution meant to jump-start the market for 3G.

Bluetooth explained
Are you still with me? I’d like to explain one more emerging technology in just a little detail before moving onto less engineering-oriented topics. Bluetooth, which was widely advertised as just over the horizon at last year’s fall Comdex, is finally rolling out to consumers this month. The idea behind Bluetooth is both easy to explain and very powerful: it is a wireless protocol meant to replace all cables used in computers and other electronic gear. Think about the last time you had to untangle cables, or deal with a broken connector, or had a cable that was just a little too short, or tripped over a cable and sent something crashing to the floor, or had to rewire your stereo components, or lost your telephone headset in the middle of a call because the wire got wrapped around a doorknob as you were walking by. Get the idea?

Bluetooth is designed to use a very low power signal to connect devices that are within ten meters of each other. Because it is meant as a cable replacement, the devices are designed to be as small, cheap and easy to use as possible. The target price for the chipset needed to make any device understand Bluetooth is around $10. Actual Bluetooth transceivers cost more than that, but eventually the cost of embedding Bluetooth into a laptop, stereo, mobile phone or other device should drop to the $10 range. Right now, Bluetooth PC cards for laptops and desktops are in the neighborhood of $100.

The first likely use of Bluetooth will be to connect a mobile phone to its headset. Bluetooth-enabled phones and headsets are being shown at this Comdex and should be available before the end of the year for a slight premium over conventional phone and headset prices. One effect that is widely predicted here is that people will start walking around with earpieces permanently attached to their ears, cell phone either in a briefcase or purse or on a belt. Ignoring the obvious benefits, this will at least remove the social stigma from people who wander the streets talking to themselves.

Now for a little technical explanation. Bluetooth is a packet data protocol, much like 3G wireless or the Internet. It uses what’s called a frequency hopping technique, in which successive packets are sent on different frequencies. This reduces interference between Bluetooth devices and makes it difficult to monitor or tap a transmission from anywhere other than the intended sender and receiver. Each Bluetooth device can communicate with up to eight others at once, serving simultaneously as the master or controlling device in some communications and as the slave device in others. All Bluetooth devices are certified by the Bluetooth Industry Association to guarantee that they will interoperate with all other Bluetooth devices.

Unlike mobile phones and broadcast radio and TV, Bluetooth uses a part of the radio spectrum that is not assigned to any specific purpose, 2.4 gigahertz. Because this frequency range is unassigned, many devices share it, including cordless phone handsets, baby monitors, microwave ovens and the 802.11 wireless networks that are becoming popular at the same time as Bluetooth. Because the frequencies Bluetooth uses are so crowded, it was designed specifically to lock onto other Bluetooth signals and reject interference from other sources broadcasting at 2.4 gig. While a number of people have expressed concerns about interference, particularly between Bluetooth and 802.11 networking running in the same PC, the Bluetooth engineers claimed that the protocol rejects interference well enough to make this a non-issue. The biggest problem is apparently interference with cordless phones, so be aware of this possibility when you take your Bluetooth laptop home.

Bluetooth connections are based on profiles. Each profile is supposed to describe the kind of cable that is being replaced. There are currently eight profiles supported in the first version of the Bluetooth standard. Some of them are RS232 for interconnecting serial devices, audio for connecting headsets and speakers to phones and audio components, video for connecting computers to displays and projectors, and infrared to replace the IR ports on many laptops, PDAs and cell phones.

One profile uses Bluetooth to do something a cable can’t. A Bluetooth-enabled cell phone can be configured to talk to a Bluetooth-enabled wireline phone. Bring the cell phone within range of the wireline phone’s base station, and it becomes a cordless handset and communicates through the land line, allowing calls to be made much less expensively and without airtime charges. Step away from the base station and it communicates through the cellular network.

The designers of Bluetooth intended it to be a hardware-level protocol, literally a cable in every sense except the physical. So each profile supports the same data transfer protocols as the cable it replaces. The serial profile supports RS-232, Infrared supports IrDA, video carries VGA, and so on. The exceptions to this are the audio profiles, since audio cables carried no digital data and required no protocols.

Second generation Bluetooth will include new profiles, including ones for printers and digital cameras. These will not only eliminate cables connecting a PC to a printer or camera, but will allow digital pictures to be printed directly from a camera, eliminating the need to download to a PC. It will also support programmable functions. Imagine approaching your Bluetooth-enabled VCR and having it communicate with your mobile phone, using WAP to display a menu of VCR functions on the phone’s screen. Your phone handset then becomes your VCR remote. This is the kind of hybrid application that Bluetooth will make possible as it is installed in increasing numbers of devices.

Timing modulation: is the bandwidth problem solved?
If you still haven’t gotten enough of radio, please stay with me for one more piece of technology. This came out of a “future watch” session and describes a new data transmission protocol, which the vendor, Time Domain, Inc., hopes will deal with the problem of radio spectrum limitations once and for all. Unlike conventional transmission, which puts data on a continuous carrier signal, Timing Modulation, also known as Ultra-Wideband, uses short bursts of data in a preset, timed pattern to broadcast information. Pulses are sent at a rate of forty million per second, each one carrying a single bit of data. The receiving device knows what the pattern of the sending device will be, waits for pulses to show up in exactly that pattern, and locks onto the signal once it is found. It then looks for slight variations in the timing of the pulses. How slight? A variation of one picosecond – a billionth of a second – ahead of the expected timing indicates a one bit, while a picosecond behind indicates a zero.

This method offers a number of advantages over conventional transmission, some of them quite surprising. Because it uses bursts of data, the signal is immune to multipath interference, a kind of interference that happens when the same signal arrives at the receiver at different times after bouncing off objects along the way. The short duration of the signal means that massive numbers of devices can share the same frequencies, and the pattern-based identity of the signal makes it very resistant to interference. Because of this, the vendor has designed his devices to operate in the 300 megahertz band.

300 megahertz is an interesting place in the radio spectrum. The Federal Communications Commission calls that the Class 15 band and uses it to dump unwanted radio emissions from virtually any electronic device. If you build a laptop that generates radio “noise” (and they all do), the FCC makes you tune that noise into the 300 MHz band before approving your laptop for sale. Same for TVs, radios, automobiles, and just about anything else with electronic circuitry. Radio guys call it the garbage band (I heard them) because no useful signal can make it through the noise – at least not without using so much broadcast power that they screw up the devices that are emitting the garbage.

Timing Modulation addresses these problems. It is resistant enough to interference to operate in the band. It doesn’t need much power to have its pulses heard through the steady din of the garbage signals. Also, a device at 300 MHz intrinsically requires less power to broadcast than one running at the higher frequencies that all other devices use. So the transmitter runs at a puny 50 microwatts, less than a tenth of a percent of the power required by a cell phone. A side benefit of this is dramatically better battery life for self-powered devices. And since bandwidth is the inverse of wavelength, there’s a lot more of it at a low frequency like 300 MHz than there is way up at 2.4 gig, even before you take the more efficient use of the bandwidth into account. Maybe enough bandwidth to run unregulated.

If that’s not enough to make you like Timing Modulation, try this: The nature of the signal makes accurate location services almost trivial. How accurate? The presenter claimed to be able to locate their device to within three inches in all dimensions. He suggested - - tongue in cheek, I hope - - that vendors could send you an instant coupon in the supermarket if your device senses that you’re reaching for a competitor’s product. The location capabilities can also be used to implement a “personal radar”, a security shield that can notify you if anyone moves within it.

There are some obstacles to adoption of this technology. Right now it’s illegal to sell products that broadcast an actual signal in the 300 MHz band. The president of Time Domain modestly calls this a “barrier to entry”, and he is negotiating with the FCC to have the restriction lifted. If he can accomplish this, Timing Modulation devices may show up on the market soon after.

Want to read more? The company’s website is at http://www.timedomain.com or http://www.time-domain.com . Both addresses appear in their literature.

Wireless security
Still here? Good. We’re done talking about radio. Before I sign off, I want to cover one more topic, wireless security. The session on this subject covered both security of transactions via encryption and authentication, and security of devices via personalization.

Transaction security
“A transaction on a wired network is inherently more secure than a wireless transaction.” It’s a commonly held belief that plugging things into the wall makes them more secure, but it ain’t necessarily so. The same tools available for Internet security - - notably SSL encryption and public/private key authentication - - are available to wireless devices. The level of transaction determines the level of security desired. For low level uses of the net, such as browsing and retail purchasing, privacy is the primary motivation for security. For vertical industries and corporate intranets, authentication and the ability to make actions legally binding are the motivators. (For people who keep track of buzzwords, the act of making sure an online transaction is legally binding is called non-repudiation.) Wireless security is much easier to implement in Europe and Asia, where equipment and networks are more standardized, than in the US.

Even in the US, encryption or key-based security can be implemented relatively easily once the transaction enters the wired network. Over the air, security ranges from built-in 40-bit encryption on some devices down to the limited security inherent in the difficulty of tapping a wireless data stream. In all areas, the same constraints that limit other kinds of services on mobile devices - - cost, battery life, speed and bandwidth - - limit the use of end-to-end security for the time being.

Device security
One presenter spoke at length about the use of SIM cards, which carry personal identity information in so-called “smart phones.” SIM cards are widely used in Europe and are available on PCS cell phones in North America. The card generally carries both identity information and operating system information for the phone. A SIM can be used to provide secure access to network-based applications and data based on the subscriber’s identity, and allows the phone to be used as a transaction center and electronic wallet. With 400 million SIM-based 2G phones in use in Europe, phone-based fraud has been virtually eliminated. And since a subscriber’s personal information is entirely on the SIM card, plugging the SIM into a new phone instantly personalizes it for the subscriber with no intervention needed from the service provider. This makes the phone itself a commodity that can be bought at any store. Service providers no longer need to sell and support the devices, activities that generally lose them money.

Certificate-based security
A representative of Verisign discussed the use of digital certificates and trusted signing authorities for wireless transactions. They currently support server certificates, and will support client certificates on Palm, Blackberry, Windows CE and SIM-based phones in first quarter of 2001. Certificates can also be used to sign application code downloaded to the phone, much as they are used in Web browsers. This is important because the ability to program a phone remotely opens the door to hackers who want to create viruses to damage or otherwise abuse cell phones. Japan’s i-Mode phones, in particular, support Java and are especially susceptible to this kind of attack if not protected by code signing.

A few words about the Luxor Hotel
Once again, I’ve run out of time to list the booths I visited yesterday and today. There have been only a few, but they’ve been interesting - - rollup keyboards and 62” TV screens you balance on your nose only begin the list. I’ll try to do a greatest hits list as part of tomorrow’s report.

Finally, I'd like to describe the hotel I'm in. Luxor looks like The Great Pyramid would have, if King Tut had been able to get his hands on about an acre of smoked glass. It's topped by a light that is allegedly the brightest in the world, and visible from the Space Shuttle. The pyramid is an atrium 30 stories high and 600 feet per side at its base, which makes it a pretty fair jog to your room if you stay on a lower floor and turn the wrong way leaving the elevator. The elevators themselves are worth describing. With an atrium at the center and the rooms sloping away from the peak at a 60 degree angle, a conventional elevator in the center of the building would look rather odd and ugly - - not to mention how dangerous it would be for people getting off anywhere below floor 29.

So the clever builders installed four angled elevators that run up the corners of the pyramid. Press the "up" button and you head off at 30 degrees to the horizontal. With no windows to the outside, it feels like more like a subway car than an elevator. You lean to one side when the elevator starts and then back the other way when it stops.

Now I know exactly what you're thinking, and I had exactly the same thoughts right away. The answers are, yes, the emergency stairs do run straight down the corners next to the elevators, and no, they do not make this the best place in the world to bring a Slinky. Somebody must have decided that a 489-foot straight stairway (based on my memories of high school trig; I've given you all the information you need to do the math yourself) might be a little dangerous. So they offset the stairs from side to side every other floor. You can stare straight down this nearly 500 foot tunnel, but your Slinky will just go down two floors and stop on a landing, just like it would in any other building in the world. Consider it a missed opportunity.

 

Copyright © 2001.
Computer Sciences Corporation.
All rights reserved.
(spacer image) (spacer image)
(spacer image)
(spacer image) (spacer image)
lef homenews and events programs the lefbookshelf
(spacer image)
(bottom gray line)