Bluetooth Information

You saw your first Bluetooth device three years ago and were suitably amazed. Now, everyone you know has a Bluetooth-equipped mobile phone, most use it to talk to their PDA, and it has become as common in the notebook as the USB ports it may one day replace.

Bluetooth Information
You saw your first Bluetooth device three years ago and were suitably amazed. Now, everyone you know has a Bluetooth-equipped mobile phone, most use it to talk to their PDA, and it has become as common in the notebook as the USB ports it may one day replace.

Or at least that's what we thought 2002 would be like when Ericsson first announced the marvel of Bluetooth technology in 1998. Since then, the Bluetooth ride has been more akin to extreme skiing than even a roller-coaster - any ups appear to have been overshadowed by downs. The result: many observers are measuring the technology up for its coffin.

But don't make a date in your diary yet. This technology was named after a warrior king and it intends to fight for its life.

When we were young

At first, everything seemed to be going quite smoothly. Toshiba successfully demonstrated an early prototype of a Bluetooth-enabled notebook at Comdex Fall in 1998, with two devices communicating at a distance of 10m. True, production costs were estimated at a crippling $30, the transceiver was too large and Microsoft was still conspicuous in its absence from the Bluetooth Special Interest Group (SIG), but these were surely just minor blots on a bright horizon.

Unfortunately, things started to go awry in 1999. Although there were some simple telephony products - such as Ericsson's Bluetooth headset and matching phone - the first wave of IT products never materialised. At the time when finished products were allegedly meant to go on sale, late 1999, Intel was merely demonstrating wireless synchronisation between notebooks.

This whetted people's appetites for CeBIT 2000, pre-hyped to be Bluetooth's moment of triumph, but Europe's biggest IT event proved to be something of a damp squib. The closest we got to finished items were proof-of-concept devices.

Things weren't helped by Microsoft failing to integrate Bluetooth support when it released Windows XP in 2001, especially as 802.11b support was integrated, prompting some to predict the early demise of Bluetooth. Microsoft assured us that this wasn't the case, pledging continued faith in the standard and citing a lack of products as the problem. But the sum of all these troubles meant that last year, when the technology was meant to be donning its crown, it was more akin to the court jester.

Fortunately, the tide started to turn in 2002, when we finally saw a string of products rather than broken promises. An ever-increasing number of mobile phones are still being released with integrated Bluetooth, and the same goes for PDAs and notebooks. Even manufacturers like MSI and Gigabyte are flirting with the standard by integrating them into their motherboards.

It's not just IT that can benefit from Bluetooth. Companies like Ford and Chrysler have already put their money where other people's mouths are, while Audi has just had a Bluetooth headset qualified for use. After all, if you've got a Bluetooth-equipped mobile phone, all you need is a Bluetooth headset and the near-obligatory hands-free car kit is reborn.

In the world of Bluetooth, however, there's no such thing as good news all round. In late August, Conexant suspended research into the technology and stopped selling its Bluetooth-related solutions. This prompted some journalists to (again) predict the demise of Bluetooth, perhaps not realising that the company only had a tiny market share in Bluetooth solutions.

A more damaging blow came from a report by research firm Gartner, suggesting companies that invest in a Bluetooth-enabled mobile phone, PDA or notebook could pay an extra $70 (£45) in support costs.

And all the while, there lurks the spectre of competition - other wireless technologies that promise more for less money.

THE OTHER WAY

In the time since Bluetooth was announced to the world, a number of rival wireless technologies have been unveiled - or, in Ultra Wideband's case, dusted down and given a spring clean.

Ultra Wideband (UWB) is far from a new technology, having been first mooted in the 1950s. It doesn't rely on any one frequency, but spreads its signal across a whole swathe of frequencies. It promises transfer rates up to 500Mb/sec, rivalling USB 2. It also boasts location awareness (similar to GPS) and very low power consumption.

Low cost, low power and low data rate, ZigBee (IEEE-802.15.4) is another standard that could yet prove to be a popular alternative to Bluetooth. It lacks Bluetooth's flexibility, but this means that protocols are simpler, so they should work with similar ease to infrared. At $2 per chip, they're also likely to be cheaper than Bluetooth, which may reach $4, but won't go much lower.

However, ZigBee is yet to be ratified and it's expected to mainly appeal to the home and SoHo user. It's also slow, boasting a 10Kb/sec to 115.2Kb/sec transfer rate compared to the 400Kb/sec to 1Mb/sec of Bluetooth. This means multimedia transfer is out of the question, which is bound to put off the mobile phone community.

And Matthew Towers, managing director of IMS Research, which specialises in wireless technologies research, is unconvinced. 'ZigBee's main advantage is that it's a very low-power solution, but the speed that it offers isn't anywhere near as high as Bluetooth. If it succeeds at all, I can see it in niche applications but not really across the broad range in which Bluetooth is expected to participate,' he said.

'It's still very, very early to say what's going to happen to UWB. My perception is that it's at least three years behind Bluetooth. There are still lots of things that have to be sorted out for UWB - like will it work? And will it work at a reasonable cost? What are the technical issues? Will they be solved? Until we start to get the answers to some of those questions, we don't know if it's going to be a threat. So it's not going to be a threat in the next two or three years and, if Bluetooth hasn't established itself by that time, it's struggling anyway.'

THE CURSE OF 802.11

Bluetooth's biggest rival, in many people's eyes, is 802.11. When Bluetooth was in its early stages, people looked at this unproven technology and 802.11b and declared the latter the winner. It's easy to see why. Bluetooth has a shorter range, a smaller bandwidth and early demonstrations of the actual products were shaky to say the least. In contrast, 802.11 worked well and already had a huge market presence.

'The Bluetooth community didn't help itself by promoting Bluetooth as an all-singing, all-dancing wonder solution that would do anything for anybody,' said Towers. 'But that's changed in the last two or three years. The Bluetooth community has settled down and focused on the area where it's going to be really strong. It's highly likely that many products will include both 802.11b and Bluetooth. Notebooks and desktop PCs are classic examples of what will include both solutions.'

The problem with 802.11 comes when you try to use it out of the office. Yes, an 802.11 notebook can exchange data with an 802.11-equipped PDA, but only as a peer-to-peer network. And you're never going to see an 802.11 radio in a mobile phone, because 802.11's power demands would cripple the talk and standby time.

It seems likely that 802.11 (in all its alphabetical forms) will co-exist with Bluetooth rather than compete with it, so the way forward is to integrate the two technologies into those devices that will move between the different areas.

Notebooks are the most obvious candidates, and we're already seeing notebooks that integrate both Bluetooth and 802.11. Sony and Toshiba released such laptops during 2002, although others are holding back, citing interference problems.

We know from our own experiments that using Bluetooth simultaneously does affect the bandwidth of the 802.11 connection, but not to a catastrophic effect. And according to Glenn Collinson at Bluetooth specialist CSR, things will get better in 2003. 'Early in the year you'll see notebooks where there's a hardware handshake between the two radios, so if you're transmitting in 802.11 it will switch off Bluetooth.

'A more sophisticated version is where Bluetooth will be told by the 802.11 what frequencies are being used and it will avoid those frequencies - advanced frequency-hopping.'

Perhaps this will be the killer argument against those who believe that 802.11 and Bluetooth are rivals. After all, just because you've got an Ethernet adaptor in your PC, you still need USB slots to connect your peripherals.

SENSE OF SECURITY

Security is a massive issue these days, and wireless technologies have come under more fire than most. You need only look at 802.11b for proof. So many wireless networks leak out into the public sphere that knowledgeable notebook users can easily surf the Internet using them. And it only takes a certain amount of willpower to break into a badly secured network and access what should be confidential documents.

Bluetooth is less likely to cause security problems. For one, it's usually based on data passing between two devices, limiting the scale of leaks. What's more, before you even form a partnership with another device - which requires permission at both ends - you need to exchange a numeric passkey, much like a PIN. The only problem with this is that passkeys are hugely guessable - try 5555, 1234 or, indeed, nothing and you're likely to gain success. Apparently, the industry doesn't consider security as a big issue.

'IMS Research did a survey, and within our research we talked to the big OEMs who designed Bluetooth into their phones, their cars and their PCs, and security hasn't been highlighted as a major issue for anybody. So I can only conclude that it's not an issue,' said Towers.

But this doesn't mean that Bluetooth is impregnable. Just as with 802.11b, there's the possibility that others can intercept your data as it's being transmitted from one device to another. However, unlike 802.11, Bluetooth uses frequency-hopping to make this more difficult. As with any wireless protocol, Bluetooth also has built-in encryption. If security is required, a 128-bit random number - the link key - is used in the authentication process. If two devices share a partnership, when that partnership was set up a link key would have been defined.

There are other hurdles too, as TDK shows in its rather tongue-in-cheek guide to hacking into a Bluetooth 'conversation':

'First hang around other people's offices or follow people in the street, staying within 10m of them. Try not to look suspicious with those three supermarket trolleys of equipment, all those cables and the antenna.

'Now, tune your receiver into the 2.4GHz band, find one of 79 active radio channels. Very quickly spot some packetised data, demodulate and de-encrypt it using the 128-bit encryption key for that particular radio dialog.

'Next, do some processing and understand which radio channel the Bluetooth radio will switch to next; a clue is that it will be on one of the other 78 channels available, but keep working quickly as you've only got 625 micro seconds to do this since the start of step two.

'You got this far? Pretty good! Go back to step two and repeat. Still with us? Fantastic! Okay, you've got two packets. Since the data rate of the Bluetooth session could be running at 768Kb/sec, that equates to 0.02604 per cent based on 100-byte packets of user data per second. By the way, don't forget to check those packet-sequence numbers and CRC checks so you can postulate what order the packets go in and whether they contain real or corrupted data.

'Still got a signal? Make sure you stay within 10m. We're truly amazed you're still with us. Want a job? Okay, it's back to step two again. By the way, time for another tip - those radio packets may contain lots of different things, it may not be data at all.

'Getting tough now, breathe deeply! Okay, if it's not data, what is it? It could be voice! Try not to scream like that. It might spook the person you're following. Yes, that's right, Bluetooth can carry voice and data. Did you bring the voice-decoding software algorithm with you?'

PROBLEMS, PROBLEMS

So this is a technology that seems to have it all - it's secure, powerful and affordable. But it has a problem. Unless you happen to be technically literate to the point where computer nerds look at you nervously, getting your Bluetooth products working together seamlessly is stupidly difficult.

For example, installing a Bluetooth USB adaptor can introduce no less than five virtual COM ports to your PC. Although the manufacturer of the adaptor produces its own software to control all these interfaces, there's no guarantee that the manufacturer of your PDA or phone is handling things in the same way. Take a look at The out-of-box experience: making Bluetooth work, opposite, for details of our experiences.

Sad as it is to admit, Bluetooth is crying out for support from the most important company of all: Microsoft. It desperately needs built-in support in Windows, just as Microsoft belatedly introduced support for 802.11b in Windows XP.

Microsoft has just released software to manufacturers that will hopefully solve many of Bluetooth's problems. 'It will make it easier for manufacturers to issue a standard Bluetooth stack,' said Microsoft's Weeks. 'Before, you were reliant on manufacturers to write a good driver for the device. If manufacturers wrote to their own standards, people could get a blue screen in Windows. This release will improve interoperability.'

Microsoft admits that support has been a long time coming, but Weeks promised that the wait has been worthwhile. 'We've got some really high goals to make sure customers get a good experience. They need to know that Bluetooth is going to work. People want simple, easy devices they don't need to configure.'

The bad news is that it will be three to six months before we see the fruits of Microsoft's endeavours, in the form of updated software and drivers from manufacturers. And they will only work if Service Pack 1 is installed on your Windows XP machine.

DAGGERS AT DAWN

So Bluetooth isn't dead, but it's far from being in A1 health. The out-of-box experience is being addressed, but we must trust in Microsoft and the various manufacturers that in six months' time devices will work together seamlessly. We also need more PDAs, phones and notebooks with built-in Bluetooth.

The more distant clouds have silver linings as well. Although the industry is likely to concentrate on stability in favour of upping the specification in the near future, Bluetooth is expected to increase in speed to 3-4Mb/sec and 10Mb/sec in coming years, which will help to fight off upstarts like UWB.

But we should learn a lesson from Bluetooth's struggle. Much like the British tabloids build up celebrities just to make their fall from grace all the more glorious, don't believe all the hype about new technologies. They may promise much on paper - just as UWB and ZigBee do now - but compatibility, operating system support and cost are all major barriers.

Ultimately, if the technology is good enough, and it has enough uses, it will battle through the barriers. Bluetooth has now done exactly that and perhaps - just perhaps - in a couple of years' time it will actually be in every notebook, every PDA and every phone, as we hoped it would be by now.

Bluetooth: dead or alive?



Local Articles
Telecommunications
Home