George, thanks for the information. I won't have time to read it just yet, but here's how I understand the various systems (ref various IEEE journals and the IEEE 802.11 handbook):
Bluetooth is competing head-to-head against the IEEE 802.11b standard as a protocol for wireless LANs. Both operate in the 2.4 GHz ISM band, Bluetooth using frequency-hopping and 802.11b using direct-sequence. Bluetooth raw data rates are 1 Mbps vs 11 Mbps for 802.11b, the difference being due to the modulation rates (equivalently, the signal bandwidth). (N.B. the 802.11b standard will dynamically vary the signaling by using shorter code sequences as the radio channel deteriorates due to range or multiple interferers, dropping from a max of 11 mbps to 5.5 and then to 1.)
Bluetooth currently employs much lower power, limiting range to 10 meters (using less than 1 milliwatt radiated power, by design) vs 30-100 meters for 802.11b (limited by FCC regulations to 1 watt in the U.S., and 100 milliwatts in Japan and Europe). Sometimes it is quoted that the 802.11b system can allow roaming up to 1000 meters away, but that would be line-of-sight unimpeded ('free space'), not the indoor environment it was designed for; because the indoor propagation losses are around 30 dB per hundred FEET, the IEEE design notebook quotes the likely usable range for this system at 100 feet. The upshot is that within a corporate campus, a given building would likely require multiple wireless LAN servers.
However, the 802.11b networks operate better and with faster data rates if the users are not at long range anyway (otherwise, one user at high rate would basically consume the entire bandwidth of the server). Both systems use much less power than a typical cellphone.
A key issue is security: Bluetooth (in its current version) is just a 'dumb' communications link protocol, and does not employ user identification or encryption in its current form. It was designed on the presumption that there is only one user's signal on the channel, and the only issue was synchronization and simple handshake protocols. To extend it to a multiuser 'piconet' application will first require another (applications) layer of software already employed in the 802.11b standard. The 802.11b standard allows a roaming user to connect to a server as if he was on an ethernet connection. However, even that standard does not employ firewells, and if a user on that network desired to connect through his corporate internet, he has no security on the link employing the 802.11b.
To come full circle, while Bluetooth *could* be employed for a small local wireless network, it currently has so many features lacking that a system designer would be better off following the IEEE 802.11b standard. The latter would allow one to wirelessly connect to someone else's LAN, but there are additional security issues. To employ Bluetooth as the AA report apparently suggests would likely be no more cost effective than using hardware already being deployed based on the IEEE 802.11b standard. (Prices will drop, as COMS, CSCO, and everyone else sees a lucrative market for such wireless 'servers.')
====================================
For followers of the thread, it is of interest to compare a WLAN system using 802.11b to a wireless system allowing roaming (such as Ricochet or the various 2G/2.5G/3G proposals). The key point is to recognize that the 802.11b systems allow the user to be wireless, but on a leash. Within a large corporate campus, it is a useful solution. But sitting in a starbucks or in the Admiral's Club at the airport, you pay by the minute to hook up the the local wireless LAN while you sit at your table or in your comfy chair. If you want to move 100 feet away, you're off the network. When you get to your customer's place of business, and want to access your home corporate data base, the customer would have to provide you with a login and password to wirelessly connect to their 802.11b LAN, and then you still have those nagging security issues. If you are a salesman or field service rep on the move all day, this is not the solution.
Ricochet overcomes all the above obstacles: complete roaming, always 'on', inherent security built in. Worldcom is reselling the Ricochet service as part of their 'IP VPN' offering to corporate clients.
Summary: Bluetooth is designed to work at close range, to allow your cellphone, pda, or laptop to communicate with your desktop. IEEE 802.11b is useful for within a corporate or school environment, where all users are already assigned accounts, so never need to reinitiate; also, firewalls can be set up once and then not bothered with. But independently owned WLAN systems in various disparate locations offer such poor overall coverage that they are not the ultimate solution for a roaming user, and given the pricing structures that will be employed, would appear to serve the niche of someone who seldom goes wireless but just happens to want to surf the web while waiting in a starbucks, and is willing to pay the $7/hour 'nuisance fee.' (Aside: MobilStar is attempting to establish relationships with various franchises, such as the two mentioned above, and then offers a flat monthly fee; but for about the same fee you can have the Ricochet service from Metricom and not be limited to a couple of locations, but roam citywide.)
I see the WLAN systems proposed for deployment in public locations (e.g., such as Mobilstar; vs the use in a widespread corporate environment where the goal is wireless ethernet, managed by a single authority) as serving a different and more limited niche. Users who want to be online wirelessly more than a few hours per month experience much better pricing by using Ricochet, and they could then dispense with their wired connections at home if desired (or, receive 'bundled service' rates from their ISP, such as Earthlink is offering now for Ricochet; you will see much more of this now that the reselling of Ricochet is commencing in earnest, for a large footprint having solid and stable coverage).
=================================
Just some thoughts. Hope some find them useful. |
