Remember the Bluetooth buzz from last year's Comdex? The market was supposed to be invaded by wireless printers, PDAs, notebooks and pagers by now, and we were all supposed to be having fun by buying Pepsi via cell phones.
Well, today's reality is that with the exception of Toshiba Inc.'s Bluetooth PC Card, our wireless shores are still empty as we wait for the Viking hordes to disembark.
What's the hold up? Is it security, which is always a bugaboo when you're talking about wireless communications? Is it spectrum competition with 802.11? Or is it that the device manufacturers want to get it right, rather than getting it on time?
These questions were put to a panel of Bluetooth gurus, which was moderated by Herb Bethoney, director of eWEEK's Corporate Partners program.
The panel included the authors of the recently released "Bluetooth Revealed: The Insider's Guide to an Open Specification for Global Wireless Communications." One author, Brent Miller, is a senior software engineer in IBM's Pervasive Computing division, in Research Triangle Park, N.C.; the other, Chatschik Bisdikian, is a research staff member at IBM's T.J. Watson Research Center, in Hawthorne, N.Y., and technical liaison with the IEEE's 802.15 working group. Also on the panel were eWEEK Labs' Brooks and Technology Editor Peter Coffee.
Bethoney: Where is Bluetooth right now, and where will it be in the near future?
Coffee: Bluetooth is now suffering from a slight case of stale promise. The concerns arising now have to do with the contention for the use of that "kitchen-sink" band in the 2.4GHz region and how Bluetooth and HomeRF and 802.11 are all going to play nicely with others. The IEEE is now trying to address that under the umbrella of the 802.15 effort.
They're quite optimistic that the capabilities that are envisioned by the Bluetooth spec and the pragmatic attitude of "Let's do what we can and get it out there and see what people really want to do with it" will result in delivery, probably within the next three to six months, of a reasonable diver sity of products; and that, as people make use of it, the demands will dictate the subsequent evolution of a specification in areas that are, right now, explicitly cited in the specs as areas where they have not yet attempted to specify mechanisms and behaviors.
Miller: We've had a stable specification for over a year, and we are starting to see the first products. Perhaps some of the prototypes and demonstrations led many to believe that products would be generally available [earlier].
A lot of manufacturers, including IBM, are trying to ensure that the first experiences with the technology by end users are good ones, and that might account for some of the delay. People are trying to make sure that the products they put out are going to work well together and have had some interoperability testing behind them.
I think 2001 is when we're going to start seeing the ramp-up toward what has been projected and seeing the diversity of devicesâ€"and to see things a little more widely deployed and in a little more common usage.
Bethoney: Where will the networking capability of Bluetooth fit into the networking portion of wireless, where the 802.11B spec already lives? Will Bluetooth stay strictly in the scope of PAN [personal area networks]?
Coffee: I would hope that Bluetooth would function as an aid to the larger-scale networkâ€"for example, by using a Bluetooth device to give your wired network constant information on where physically you're located so that information can follow you and can be near you when you want it. That would be a tremendously valuable synergy between what Bluetooth is designed and specified to do and what the 802.11 or other networks are designed to do.
It would probably be a mistake for Bluetooth to try to spread its umbrella too far and go beyond PAN to try to become all things.
Miller: The design points of 802.11 and Bluetooth are optimized for particular usage. In the case of 802.11, that is, in fact, the traditional LAN/Ethernet-style networking. In the case of Bluetooth, it is optimized for PAN.
Which isn't to say that there isn't some overlap; Bluetooth can do some things that begin to resemble networking, but they're not the primary usage case, which starts with cable-replacement scenarios.
Bethoney: The IEEE also has a wireless PAN initiative, the 802.15. They're going to specify the media access layer as well as the physical layer. Will Bluetooth go beyond that?
Miller: Bluetooth is going to be primarily in the PAN. There are efforts under way to figure out how we can do some more things that start to resemble more classical networking in ad hoc networking situations, but that's not 802.11.
Anyway, they are complementary; Bluetooth was designed to do PAN kinds of scenarios. Similarly, 802.11 was designed for classical LANs. While you could argue that perhaps it might have applicability in some of the PAN scenarios, the same thing applies; that isn't what it was designed for. It wasn't designed for mobile devices and low power consumption.
Bisdikian: The 802.15.2 task group within 802.15 deals with issues like coexistence between 802 wireless technologies and, primarily, between 802.15.1 and 802.11, or, in other words, between Bluetooth and 802.11.
The objective of the task group is to develop guidelines, or recommended practices, for the good behavior of the two technologies, if any coexistence problems are found to exist.
Clearly, since both technologies share the same bandwidth and become more and more popular on portable devices, one would need to study in full the implications of the two technologies to each other. I believe that both solutions could coexist reasonably well, but I will wait until the technical community and the coexistence group study the issue and develop their guidelines to see what makes sense to do.
Bethoney: Does the 802.15 committee plan to have a standard by this time next year?
Bisdikian: Yes, and they have a program for that, but we never know how things will turn out within 802.15. Taking .11 as an example, it took many, many years to develop the .11 specification.
Second, with respect to the coexistence group in .15, they're not trying to have a standard. The .15.2 group is not going to develop a standard for coexistence. It's going to develop guidelines for coexistence for the developers to follow. There's not necessarily going to be a standard coming out of it.
Bethoney: What about working group .3 in the 802.15 group?
Bisdikian: The .3 group is focusing on so-called high-rate radio. Right now, they're in the very early stages of investigating various proposals. There is no convergence yet whatsoever. At the current state, there exists a number of proposals, and the group is developing evaluation criteria for comparing the proposals.
The objective would be to build consensus for the proposals based on criteria of how each performs and under which conditions. From this consensus, the next step would be to start working on the development of the high-rate solution for the 802.15.3 group.
Bethoney: How well will Bluetooth security work when you get disparate devices together?
Coffee: On paper, the Bluetooth security is quite robust. There's encryption at the link layer. There are mechanisms for relatively robust authentication of devices. With some performance enhancements, the devices that communicate with each other routinely can retain key information so that the overhead of the security mechanism is not unreasonable.
The technology is probably suf ficiently proven that it will be implemented with a fairly high degree of correctness, and people will find the combination of security and per formance to be a well-chosen trade-off.
Bethoney: Is that security optional, or is that something that's on automatically?
Coffee: The answer is yes; that is to say, there are some security mechanisms that are intrinsic so that links are encrypted between, say, fraternally close devices.
There are also degrees of security that can be chosen at the application level, and, of course, application developers could decide whether to have additional protecting layers amen able to user control.
So, we have a hierarchy of options here, but even in the minimal default case with no user intervention required, there's still an entirely adequate degree of security for all but the more paranoid user.
Miller: Bluetooth has several different security mechanisms in place. The first relates to the fact that Bluetooth works on proximity networking. Just because any two random devices come into proximity doesn't mean they will immediately begin to communicate with each other. They will, based on the application and/or user configuration, decide which devices they do want to talk with. There's the concept of what are called trusted devices, particularly for your own PAN devices.
Frequency hopping, by its very nature, provides one degree of security, at least from eavesdropping, but beyond that, the SIG [special interest group] has specified encryption for the over-the-air traffic: device authentication; authorization for services that one wants to access from another device; and, in particular, in the profiles. The profiles, which define sort of a formalization of the usage models, have requirements and/or recommendations of the degree of security that is appropriate for a given usage case, whether that's synchronization or file transfer or what have you, including, in some cases, consulting the user for things like file transfer.
The goal has been to make Bluetooth as secure as a cable. Many people believe that there is sufficient security mechanism built in or enabled through applications that we are already at an acceptable level of secu rity.
Coffee: The point that Bluetooth isn't a promiscuous connectivity scheme is particularly important to note. Devices can choose whether to be visible to neighboring devices or, in essence, to lie low until they're specifically interrogated by someone who knows what to ask for.
The mechanism is not inherently insecure, as you might think it would be when you hear that it's based on devices discovering each other. A device can choose whether or not it's going to be discoverable.
Bisdikian: Also, Bluetooth security allows the concept of trusted relationships among your personal devices. You may introduce each of your devices to each other and allow them to connect, subsequently. Or, of course, you can allow various degrees of configurable security levels. But this would be left to the various application developers and product designers to address.
Bethoney: What about security in the products that have come out so far?
Brooks: In reviewing the Toshiba solution, I was impressed with their pretty high level of granularity in setting up security relationships. There are a lot of options as to whether you are detectableâ€"the sort of trust relationships you set up with specific devices.
I don't doubt that Bluetooth is as secure as a cable, but the difference is that it will be easier to make a connection between two devices.
If someone walks up to you and plugs a cable into your laptop, you know that something is happening. This is where we get into some possible security holes; we get into the social engineering aspects of it. Something pops up on your deviceâ€"an alert of some sortâ€"and just as people over instant messaging get out credit card numbers, I think just the added ease of connection between devices will open up some new possible human-based security vulnerabilities.
Bethoney: To where is the spec evolving, and what products will come from more development?
Bisdikian: If you have very simple sensors that want to measure temperature and height and you need them to run on a single battery for years, the current version of Bluetooth might not be the solution for that environment. There may be other technologies that might address ultralow-power and ultra low-cost solutions. IEEE will have a task group for such ultralow-power consumption devices.
But I strongly believe that Bluetooth is the first technology in the sense of [allowing] pervasive connectivity among people's personal devices, like your PDAs [personal digital assistants] and your cell phones and your notebooks and your pagersâ€"or whatever else personal you carry on yourself.
And the greatest advances for pervasive connectivity will come from people playing with the technology and developing fun applications that the Bluetooth community has not thought of yet.
