Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
A Gentle Introduction to Bluetooth

By HenryR in Technology
Tue Jul 16, 2002 at 03:54:56 PM EST
Tags: Technology (all tags)
Technology

The Internet is based, structurally, around pieces of wire which stretch from computer to computer across land and under the sea. This is a suitable structure for such a vast network, where links will remain static and the relationship between the computers is fixed.

However, what happens when we remove the wires? What changes can we make in the way we think about the permanency of networks, and the modes of use of the devices involved? By taking away physical constraints, what do we gain?


ADVERTISEMENT
Sponsor: rusty
This space intentionally left blank
...because it's waiting for your ad. So why are you still reading this? Come on, get going. Read the story, and then get an ad. Alright stop it. I'm not going to say anything else. Now you're just being silly. STOP LOOKING AT ME! I'm done!
comments (24)
active | buy ad
ADVERTISEMENT

Attempting to answer some of those questions is the Bluetooth standard, a specification for a wireless communication protocol.

Bluetooth is not the first wireless protocol - I don't know what technology lays claim to that title. I do know that there are contemporary technologies such as the 802.11* protocols, and the IrDA specification. As is ever the case, all of these protocols have advantages and disadvantages, and it is not my desire to enumerate them here (except when they serve to make a point about Bluetooth). There is no clear winner, but Bluetooth is beginning to creep in to consumer devices and occupy positions previously held by infra-red ports.

This brings us to an interesting and important point early on - what kinds of devices and networks is Bluetooth suitable for? Always keep in mind that we need to know what Bluetooth offers us that conventional wires do not. Without wanting to put the cart before the horse in terms of explanation, because we have no physical constraints we can move a Bluetooth device around, and possibly between different networks, at least in theory. To make this practical, Bluetooth devices should probably be lightweight and portable, so we are thinking about laptops and mobile 'phones, rather than large servers or even desktops. As such, this begins to place constraints on the operation of Bluetooth modules with regards to power consumption, weight and cost.

Some History

Version 1.0 of the Bluetooth specification appeared in 1999. We are now up to version 1.1 (a 10% improvement!) which was released in February of 2001. However, the origins of the technology can be traced back to a study undertaken by Ericsson in 1994 to examine alternatives to wire-based links from mobile 'phones to their accessories. From this study came the original Bluetooth specification. Prior to the specification's release came the formation of the Bluetooth Special Interest Group (SIG) in 1998. The Bluetooth SIG is made up of many companies who are promoting, adopting or actively contributing to the development of Bluetooth. The big names are there: Microsoft, Nokia, IBM, Intel, Motorola, of course Ericsson and more. You can find the full list at the SIG site. The SIG takes care of the interests of Bluetooth, from the development of the technology to protection of the Bluetooth trade mark. Joining the SIG is impossible to do at an individual level: your company must join and sign various agreements before you are allowed to join. However, since all the specifications are available unless you have something to actively contribute there is little need for full SIG membership.

The Technology:

So far, so much bureaucracy. What are the technical underpinnings of Bluetooth that give it its worth?

Bluetooth uses radio waves as the transport medium. This immediately confers some advantages - particularly the fact that line-of-sight is not required to establish a connection (in direct contrast to IrDA). The radio waves operate in the unlicensed 2.4GHz Industrial, Scientific and Medical (ISM) frequency band. Since this band is unlicensed other devices may be using the same set of frequencies. Clearly this would cause interference as each device tried to communicate with each other on the same frequency. We don't get this problem with wires because each wire represents exactly one link between two devices, and so there is no competition for the medium.

Bluetooth gets around this by frequency hopping. That is, the exact frequency used to communicate between two devices is altered every time slot (typically every 625 microseconds). There are typically 79 different frequencies to choose from. The next frequency in sequence to use is determined by a pseudo random function, with an initial phase determined by the unique address of one of the devices taking part in a communication. This ensures that any effects of interference are short lived. Obviously the chances of collision get higher the more devices you have talking over the same airwaves. However, such a volume of devices is unlikely. (Frequency hopping also provides security benefits: the US Army considers a frequency hopping scheme of 79 frequencies secure in itself. However, Bluetooth security is considerably more involved than this, and may be discussed in a future article).

The other feature of Bluetooth that prevents interference is the limited range of the radios. There are three classes of Bluetooth radio: Class 1, with a range of 100m, has a power consumption of 100 times a Class 3 radio, which has a range of 10m. Most 'phones and PDAs will have Class 3 radios, but static access points (which perhaps act as links into another network) may be more powerful. We can see that the focus of Bluetooth is local communication between devices - somewhat managerially termed `Personal Area Networking', or PAN. Usage scenarios typically focus on synchronising contacts between a laptop and a 'phone, or some similar application. Certainly Bluetooth is not going to be used to build a world-wide network. As an aside at this point, note the low power consumption of Class 3 radios: 1mW. Bluetooth is designed to have tiny power requirements, and the specification backs this up by allowing radios to be powered down when they are not active.

Since the devices involved in a Bluetooth network are typically small and portable, we noted earlier that they may be moved around. This allows for the construction of ad-hoc networks. These are loosely defined, transient collections of computers that may only require a single communication before they are finished, and may move on. Topologically, Bluetooth networks are trees, with an arc from a node to its child representing a master-slave relationship. This relationship is very important for Bluetooth, as it defines who makes what requests at what time, as well as arbitrating things like which address is used to seed the frequency hopping algorithm. Each tree is called a `piconet'. There are restrictions on the number of children, as well as the maximum number of devices in a piconet. A key feature of Bluetooth is that these piconets may be established very fluidly, with each device granted a certain amount of autonomy in the connections it wishes to establish. Of course, the final decision may be routed back to the user, and this is something it is important to emphasise: one of the main concerns of lay-users I have talked to about Bluetooth is that devices may share information without the user's consent. Security features, and good software design should hopefully prevent this from happening. Hopefully.

Applications:

There are many, many applications of Bluetooth technology. Little, convenient things like being able to use your mobile as a modem without taking it out of your bag, printing pictures from your desktop (or even your 'phone) without having to plug in and more. The advantage of Bluetooth is that now small devices can be endowed with cheap, efficient, fast network connectivity.

Some Editorial Conjecture:

Version 1.0 of Bluetooth was not a great public success. Stories of incompatability and high module prices put off vendors. Version 1.1 is said to be a great improvement, especially for compatability. However, Bluetooth still lacks a `poster-child' application to really bring into the public view. Perhaps it can survive without one - IrDA has seen reasonable consumer-space success without being associated with any one application. Certainly Bluetooth modules are appearing in more mobiles (Nokia's 6310, and Sony-Ericsson's T68 for two), and Compaq/HP's Ipaq now has Bluetooth as standard. But Bluetooth is really built around the idea of services - where devices can seek out the nearest service provider, and make use of them. This necessarily restricts us, at least initially, to well known applications - headsets, printers, LAN connections, since in order for services to become ubiquitous they require some level of standardisation. I think it is a similar problem that is facing Web Services - dynamic discovery is a fine thing, but we need to know what we are looking for first. As such we may be left with a chicken-and-egg scenario, where the more innovative services are left waiting on widespread adoption of Bluetooth, which itself must rely on innovation.

Coming Up:

If people are interested: the Bluetooth stack, and how things like TCP/IP are layered, Bluetooth security and a comparison of SDP and DISCO/UDDI/SOAP.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o at the SIG site
o specificat ions are available
o Also by HenryR


Display: Sort:
A Gentle Introduction to Bluetooth | 42 comments (29 topical, 13 editorial, 0 hidden)
First wireless communications protocol? (1.00 / 1) (#11)
by haflinger on Tue Jul 16, 2002 at 09:39:31 AM EST

That would probably be amplitude modulation.

Kids these days... :)

Did people from the future send George Carlin back in time to save rusty and K5? - leviramsey

Heh... (none / 0) (#12)
by HenryR on Tue Jul 16, 2002 at 09:41:51 AM EST

Yeah I actually thought about saying that, but didn't want to engender arguments about what constitutes a protocol.

Henry

----------------------------
http://www.desiderate.co.uk
[ Parent ]

Modulation does not a protocol make (none / 0) (#16)
by ip4noman on Tue Jul 16, 2002 at 11:54:59 AM EST

There are several radio frequency modulation methods which predate AM (albeit not very good ones!), most notably, continuous wave (still popular today amongst amatuer radio operators), and the broad-spectrum Spark-Gap generators..

But modulation does not a digitial communications protocol make. While modulation of some carrier seems to be required, it needs quite a bit more to make it actually useable: aspects which perhaps address reliability (did some part of the message arrive?), encryption, fragmentation of large messages across a medium of smaller MTU, and reassembly of these fragements, data integrity verification, etc. etc.

My vote for the earliest true wireless protocal would be TNC-1, developed for packet radio.

--
Breaking Blue / Cognitive Liberty Airwaves
[ Parent ]
Protocols don't have to be digital. (2.50 / 2) (#17)
by haflinger on Tue Jul 16, 2002 at 12:01:59 PM EST

Otherwise, we wouldn't say digital communications. Analog communications use protocols as well.

I'd forgotten about continuous wave. AM's pretty old.

You're probably right about TNC-1 being the oldest digital wireless protocol, though.

Did people from the future send George Carlin back in time to save rusty and K5? - leviramsey
[ Parent ]

Spark gap (4.50 / 2) (#23)
by wiml on Tue Jul 16, 2002 at 04:04:59 PM EST

would be a wideband system using either on-off-keying or pulse-position-modulation, depending on how you look at it. (Even the early spark gaps built by Marconi used a tank circuit for tuning, so I won't call it ultra-wideband. ;-) )

[ Parent ]
cool (3.28 / 7) (#15)
by pb on Tue Jul 16, 2002 at 11:31:19 AM EST

In my senior year of college, I was on a team that tested what sorts of interference you can expect with Bluetooth and 802.11b, and since we didn't sign anything, I guess I can tell you a bit about it...

We were testing 10dBm and 100dBm (prototype) Bluetooth cards (and 802.11b) in laptops in different environments and ranges.  In a crowded office environment, you might get interference from some cell phones, or anything that interferes with your line of sight (like office partitions, walls, etc.).  Regular microwaves didn't affect anything, however.

For cable-replacement and really short-range stuff, Bluetooth should be great.  For much past that, it's time to look at 802.11b.

Also, when I analyzed ping times in Bluetooth, I could see a lot of variation that was likely caused by the frequency hopping, but I don't think that would be an issue for most applications, although it is interesting to look at.
---
"See what the drooling, ravening, flesh-eating hordes^W^W^W^WKuro5hin.org readers have to say."
-- pwhysall

Been waiting for Bluetooth for ages... (2.00 / 1) (#18)
by Talez on Tue Jul 16, 2002 at 12:05:14 PM EST

Once its out in a proper form I'll be putting the access points in every room to make some kind of huge piconet around the house. Of course I'll be keeping my good old Cat5E 100Base-T for connections between PCs but for wireless internet browsing around the house, Bluetooth should do everything I need without the setup required for 802.11b networks.

Si in Googlis non est, ergo non est
Hear, hear! (5.00 / 1) (#19)
by jmzero on Tue Jul 16, 2002 at 12:28:09 PM EST

Right now I've got no solution for my wireless internet browsing around the house needs.  Why the Hell should I have to setup 802.11b just cause I want to read User Friendly on my way from the bedroom computer to the fridge computer?  

There is currently 75(!) square feet of my house from which there's no reasonable way to access the internet.  Sure I can read pages offline on my laptop - but what if I click on an interesting link while I'm in the stairwell (where I having a hard time wiring a good accessible jack).

For Hell's sake, when is Bluetooth going to be ready!
.
"Let's not stir that bag of worms." - my lovely wife
[ Parent ]

Portable net browser for around the house... (4.50 / 2) (#21)
by Talez on Tue Jul 16, 2002 at 12:42:31 PM EST

No bigger than an A4 piece of paper running at a resolution of 800 x 600. Runs on a bluetooth connection, low power touch-sensitive LCD and a rechargable Li-ion cell. Features limited handwriting recognition, decent web browser and limited media capabilities.

If only I could buy one or find the hardware to build it. I'd be there and have my house a bluetooth paradise in a hearbeat.

Si in Googlis non est, ergo non est
[ Parent ]

Uh... (4.00 / 3) (#20)
by trhurler on Tue Jul 16, 2002 at 12:41:34 PM EST

First of all, the military may at one time have considered a 79 fixed frequency hopper to be secure, but they don't now. Any data classified above secret must be encrypted if it is to be transmitted in any electronic form, and secret or confidential data typically has to be encrypted too, though I think they've approved the use of commercial encryption for that purpose. They might allow some form of frequency hopper to be used with plaintext if an encryption unit failed, but probably the hopper in question would be capable of reassigning its hop frequencies on the fly as part of the "conversation," and almost certainly the data would also be interspersed at prechosen but seemingly random intervals with noise to fool automated reassembly hardware. Bluetooth does none of that.

Second, I've yet to see any compelling use for this. There are other wireless protocols that operate at short range and can be used in a variety of devices, and notice that even people who would theoretically benefit simply don't use them, save a few oddballs who do it for the "cool factor."

--
'God dammit, your posts make me hard.' --LilDebbie

A common misconception about Bluetooth (5.00 / 12) (#24)
by wiml on Tue Jul 16, 2002 at 04:40:57 PM EST

is to think of Bluetooth as a wireless networking method. It's not. It's a wireless peripheral attachment method. There is a fine distinction here which may not be obvious when you first think about it, but those two realms are actually quite different.

For analogy's sake, consider Bluetooth to be the wireless equivalent of USB, and 802.11 to be the wireless equivalent of Ethernet. Think about the differences between USB and Ethernet:

  • USB allows one master (the host) to talk to a number of nearby slaves. Slaves can be very simple if necessary — like a mouse. They can't be very far away. They can't communicate independently of the host. USb is a fairly complex, multilayered protocol. For many classes of devices, the USB standard specifies the entire protocol for the device; you typically don't run other protocols layered over USB.
  • Ethernet, on the other hand, is a way to connect a number of peer systems together; there isn't an "ethernet master" and "ethernet slave". It can go longer distances. The Ethernet protocol is quite simple, designed for flexibility and performance and to be used to support a stack of higher level protocols. It's rare to communicate using raw Ethernet frames.
You can buy devices which will allow you to network machines using their USB ports. And you can buy peripherals which communicate with their host over Ethernet. But in both cases you're working against the protocol's design, and you pay for it with reduced elegance, with performance overhead, and with inflexible, kludgy, annoying workarounds. USB and Ethernet are solving different problems.

The relationship between Bluetooth and 802.11 is very similar. It's not as obvious to the casual user, since you don't see your wireless hardware; it's hidden inside your devices. But the difference is there.

My machines all have both Ethernet and USB. I'm not about to rewire my home or office LAN using USB. And I'm not going to get an Ethernet-based keyboard or mouse. Likewise, I expect that in the future the typical laptop will have both 802.11 and Bluetooth. (Possibly integrated into the same RF chipset, but that's a technical detail.) You'll use 802.11 to talk to the network or the guy across the aisle, and you'll use Bluetooth to talk to your headset and your phone and your mouse.

networking vs. peripheral attachment (5.00 / 1) (#38)
by davka on Thu Jul 25, 2002 at 12:45:46 PM EST

I don't view the distinction you are making as essential. It is true that Ethernet itself does not distinguish between client and server, but in all applications in the real world the higher levels do.

Bluetooth is not just a transport, like Ethernet, but a complete protocol stack, therefore its design could integrate the master-slave relation at the transport level. I don't see why a designed-from-scratch network stack couldn't do the same.

That historical accidents caused a specific combination of protocols that happens to be popular today to be built with different assumptions is neither here nor there.

[ Parent ]

Where are my toys? (4.00 / 3) (#25)
by SpaceCoyote on Tue Jul 16, 2002 at 05:29:22 PM EST

Well, I brought home my shiny new Toshiba laptop with built-in Bluetooth a while back, and still haven't found a use for it. The most obvious application, keyboards and mice, don't seem to exist any any form that I can go out and buy right now. Immagine the convenience of being able to set your laptop down next to your bluetooth-enabled keyboard and mouse and just start working. No more dongles, no more wires. That's what I was promised. Instead I get a feature I paid for and am not using.
___ Cogito cogito, ergo cogito sum.
It's like USB way back when (5.00 / 1) (#32)
by enry on Wed Jul 17, 2002 at 08:49:00 AM EST

When USB first came out, all the PCs shipping had USB ports, but the USB devices were few and far-between (crappy MS OS support didn't help either).

Now we've got USB devices all over the place, and you're probably glad that your laptop has USB ports.  Hopefully in a few years you'll say the same about bluetooth.


[ Parent ]

Where are my toys? (none / 0) (#39)
by davka on Thu Jul 25, 2002 at 12:57:40 PM EST

Bluetooth had really unlucky timing. Just when it was on the verge of a breakthrough with lots of nifty little gadgets on the way to the market, the bubble burst. Introduction and adoption of new products were essentially frozen.

There has been some slow progress the last couple of years, but it is too early to know whether Bluetooth is going to make it or not.

[ Parent ]

OSI model (2.00 / 1) (#26)
by Cluster on Tue Jul 16, 2002 at 07:01:23 PM EST

What layers of the OSI model does Bluetooth encompass?  Just layer one?

At least level 2 ... (none / 0) (#30)
by ukryule on Wed Jul 17, 2002 at 03:34:28 AM EST

Bluetooth provides layers 1 & 2 (Physical & Link), and you can then slap IP on top of that.

However, depending on how you're using it, it provides more on top of this. There is a concept of 'profiles' for different uses (e.g. a "LAN profile" and a "File transfer profile"). Some of these profiles will go 'higher up' the OSI model.

[ Parent ]

In hardware.... (none / 0) (#31)
by HenryR on Wed Jul 17, 2002 at 04:26:39 AM EST

the Bluetooth stack is usually implemented up to the Host Controller Interface, which is roughly equivalent to layer 4 (Transport, although it overlaps a bit). On top of that, the layers are often implemented in software. The next layer up is Logical Link Control and Adaptation Protocol (L2CAP), which is responsible for providing a logical, multi-connection interface to the hardware. After that you start getting slightly more optional layers, like RFCOMM which provides a RS232-like serial interface.

Henry
----------------------------
http://www.desiderate.co.uk
[ Parent ]

Now I am a bit confused. (none / 0) (#33)
by Cluster on Wed Jul 17, 2002 at 03:15:00 PM EST

Does this mean that if I buy a Bluetooth NIC, I have to use a Bluetooth-aware/Bluetooth-compliant TCP/IP stack?  I thought the NIC was supposed to take care of these things.

[ Parent ]
From what I understand (none / 0) (#35)
by Graymalkin on Wed Jul 17, 2002 at 11:54:30 PM EST

If you have a Bluetooth card in your PC it is going to run similarly over TCP/IP as something like a USB Ethernet adapter does. The device itself provides only physical layer access. Even framing has to be done in software. A network connection over Bluetooth would act more like a SLIP connection than an Ethernet connection.

[ Parent ]
First wireless protocol (1.00 / 1) (#27)
by ghjm on Tue Jul 16, 2002 at 07:22:43 PM EST

Bluetooth is not the first wireless protocol - I don't know what technology lays claim to that title.

Perhaps this, or arguably this.

Waaay to high-tech ... (none / 0) (#29)
by ukryule on Wed Jul 17, 2002 at 03:14:14 AM EST

What about (American) Indian smoke signals?
[I guess calling the first caveman 'grunts' a protocol is going a bit far ...]

[ Parent ]
Actually (none / 0) (#37)
by ghjm on Fri Jul 19, 2002 at 11:04:35 AM EST

What about the information-conveying dances of honey bees? This is undoubtedly a "protocol" yet it probably existed before there were humans...

[ Parent ]
So What I'm Looking For Is... (5.00 / 1) (#28)
by Baldrson on Tue Jul 16, 2002 at 09:06:52 PM EST

An ad hoc wireless network system with these features:
  • NO BASE STATIONS.
  • Dynamic and automatic creation of new routes without any user intervention.
  • Firewall per user.
  • Hard encryption.
  • Accounting system to optimize routing costs.
  • Enough range to link to the next door neighbors at 10Mbps.
  • Fixed IP address (go to IPv6 if you think it necessary).
  • No special hardware -- just reprogram the existing 802.11* standard hardware.
  • Software cost no more than the hardware.
Is that too much to ask for?

-------- Empty the Cities --------


Re: So What I'm Looking For Is... (none / 0) (#34)
by srichman on Wed Jul 17, 2002 at 08:10:34 PM EST

* NO BASE STATIONS.
* Dynamic and automatic creation of new routes without any user intervention.
Well, this basically defines ad-hoc networking. Making the second point work well is an area of active research. See, for example, Grid.
* Fixed IP address (go to IPv6 if you think it necessary).
This works fine if you just want to network with other local devices (a personal area network, as the author puts it). If you want you device to talk to the rest of the world, though, you run into problems, as current hierarchical Internet routing is based on proximate addresses implying geographic proximity. But you can probably fix that problem with a bit of indirection.

[ Parent ]
Vending machines (1.00 / 1) (#36)
by Rasman on Fri Jul 19, 2002 at 06:38:35 AM EST

All you need is for Coca-cola, Ericsson, and Nokia to get together and give people a mobile phone interface to vending machines.

You're thirsty. You see a coke machine. You check your pockets. Damn, no coins! Cell phone though! You check to see if it has the "Supports Bluetooth Commerce" logo on it. On your mobile, you select "Scan for devices". One device found, "Company: Coca-Cola, Device: Vending Machine". You select a product "20oz drink, $1". You select "Authorize transaction" (perhaps you key in a PIN at this point too). You look back to the machine to see that it registers that you've paid enough for one drink. You press the drink button you want, and out pops an ice-cold beverage! You get billed on your credit card at the end of the month. Billing options may vary, but credit card would be a good start.

How's that for a poster child?

---
Brave. Daring. Fearless. Clippy - The Clothes Pin Stuntman
credit cards (none / 0) (#41)
by bolthole on Tue Jul 30, 2002 at 09:27:27 PM EST

its not that compelling. They already have vending machines with credit card swipers.

[ Parent ]
My way more secure (none / 0) (#42)
by Rasman on Tue Aug 13, 2002 at 07:30:06 AM EST

If someone steals your credit card, they could empty the vending machine of drinks. My way they'd need to know your PIN. I guess with your way you could include a PIN, but it'd be tougher for someone to see when you type it into your phone than on a keypad on the machine.

Not to mention you don't even need to touch the machine, so you could buy a drink for a girl across the room. Well...okay, so it's not that different. But as I've never seen a credit card vending machine, I'm guessing it never really caught on. I admit, it is a lot of technology to add just to get a few extra orders from changeless consumers. So much easier just to take their physical money from them. Oh well...

---
Brave. Daring. Fearless. Clippy - The Clothes Pin Stuntman
[ Parent ]
follow-ups (3.00 / 1) (#40)
by davka on Thu Jul 25, 2002 at 03:12:45 PM EST

As to possible follow-ups: I personally would be more interested in practical advice for would-be BT hackers, for example someone trying to build wiring-less wearable computers. Where to get cheap components to play with? a comparison of different Free-software BT stacks etc., rather than descriptions of the different protocols, which are already available on the SIG's site.

In that vein here is a link which seems promising about BT and Linux on laptops.

In any case thanks for an interesting and informative article.

A Gentle Introduction to Bluetooth | 42 comments (29 topical, 13 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!